It’s time to transform the Fedora devel list into something new ===============================================================
My first post on this list was over 19 years ago. (It was about Bugzilla. I was a fan!) Ever since those early days, devel list has been the heart and center of Fedora activity. Now, I hope to convince all of us here that it’s time for something different.
As it is, devel list is too much for many people to follow — people we’d like to have around. It covers many different things at once, yet also drives us towards more scattered communications. Our infamous mega-threads are not really effective for getting to community consensus, and tend to bring out the worst in us.
I propose that we transition devel list, and eventually most of our mailing lists, to Fedora Discussion (our Discourse-powered forum).
I know this is a big change, but, hear me out…
We’re missing people --------------------
A Mastodon post from long-time Fedora contributor Major Hayden got me thinking:
How do people make so much time available for mailing list discourse?
Once I ensure my team has the technical guidance they need and I work through the tasks of work that I owe other people, I take a look at the mailing list and say: "Oh my gosh, what the heck happened here?" Then the discussion goes further off the rails while I'm typing out a reply and my reply is no longer relevant.
I know many Fedora folks, old-school and new, for whom devel list is just too much. Some of it is the sheer volume, but this “off the rails” tendency is real — threads drift, get into back-and-forth debates about particular details, etc. And… some people aren’t here because — in contrast with our “Friends” foundation, it isn’t always a nice place to be (and mailing lists don’t provide many tools for moderation, except the big hammer of outright bans).
Ben Cotton recently did some basic analysis on devel list traffic over time, and there’s a clear trend: fewer people are participating, even though the number of different threads goes up. I don’t think this is because of any decline in Fedora contributors overall — I think it’s that conversations are happening elsewhere.
Big threads are … bad, actually -------------------------------
When we have something to talk about, it tends to explode into a big thread. The thing in January with FESCo’s frame pointers decision (https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org/...) is a good example of things going badly.
Most of the conversation was under the subject “Schedule for Tuesday's FESCo Meeting (2023-01-03)”, because everything started as a reply to that. That’s pretty easy to overlook. It’s possible for replies to change the subject when replying, but that can’t be done retroactively, and then isn’t consistent (and it breaks threading in Gmail, too).
Then, things got rather hostile, making it hard to have a reasonable conversation about the issues (both technical and procedural). And then, things went in circles without adding anything new.
This could have all gone a lot better.
And that’s just one example. Take a look back at any mega-thread, and you’ll find similar — and worse. When things get heated, the only way to intervene is by adding more. There are often long subthreads of two people going back and forth on tangents. Then, other conversation branches duplicate that, or refer across. Classical email tools don’t actually handle this kind of thing very well at all. In my experience, it only really works if you keep up with the conversation in almost real time, which has its own problems even when that’s possible.
We’re scattered in actual practice ----------------------------------
Devel list may be the center, but we have _hundreds_ of Fedora mailing lists. A dozen or so are reasonably active (Test, Legal, ARM…) but most are inactive or dead. Some are just meeting reminders over and over — for meetings that aren’t even active anymore. It’s easy to make but hard to _unmake_ a mailing list.
For lists that are active, the split is confusing — when should something be on the packaging list rather than devel? What happens when something is related to both Cloud and Server, or Workstation and KDE? One can post to both lists, but if someone replies and isn’t subscribed to both, the conversation gets split.
With “devel” as the main list, conversations about marketing, design, events, and so on don’t really have a central place. (The Mindshare list never really caught on.) That makes these important activities feel even more disconnected and secondary in status — and they shouldn’t be.
Many groups have actually moved away from lists to using tickets for team conversations — both those non-engineering functions and development. Design Team has a mailing list, but mostly for announcements: the work happens in tickets. Workstation largely uses their Pagure tracker. And CoreOS conversations happen almost entirely in tickets on Github.
Tickets are made for tracking specific, actionable tasks, and that kind of tracking is part of why teams use them over mailing lists — but Fedora teams use them for open development conversations too. I think that’s largely a symptom of mailing lists not being enough for what we need. The trackers have media support, editing for typos or updates, reactions for simple agreement, tagging people, and granular subscriptions. They are effectively “off-label use” mini-forums that teams can quietly move to using without the sort of conversation I expect this message to generate.
Airplane diagram, survivorship bias -----------------------------------
Since I’m posting this to devel list, I do expect a lot of push-back. Maybe I’ll be surprised and more of y’all are already with me on this and just waiting for something to happen. But overall I expect a tough crowd.
You’ve probably heard the story about bullet holes in airplanes returning from missions and the accompanying diagram — find it athere https://en.wikipedia.org/wiki/Survivorship_bias if that’s new to you.
The set of remaining regular participants on this list is naturally biased towards those for whom it is working just fine. But, core Fedora development discussion can’t be limited to that ever-shrinking group. Consider who isn’t here. The problems are real, and the trend isn’t in a good direction.
Devel List is too many different things! ----------------------------------------
We use it for Change discussion (resulting in those not-actually-so-great big threads).
We use it for introductions and onboarding — we’re usually pretty good at that, actually (but it adds to the overall load of following the list). We’re not very consistent, though.
We discuss packaging: guidelines, help on different topics, coordination on specific work. There’s unclear overlap with the packaging list, though, which is a bit confusing.
We talk about higher-level Fedora OS development topics that don’t fit anywhere else. For example, this on installer environment size: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org/... Sometimes, these are FESCo topics (which FESCo heroically struggles to keep here rather than on Pagure tickets…).
Sometimes, people talk about a particular interest in the hopes of forming a Special Interest Group. Often that does result in a new SIG launching — sometimes with a brand-new mailing list which then ends up getting only a few posts ( and a whole lot of spam years later). But this really suffers from the survivorship problem: many people who might be interested will never notice.
We have release process announcements — mostly to devel-announce, but then occasionally replies and discussion here. Blocker bugs and other test-list and QA topics are cross-posted, as part of the release cycle.
And then there are a lot of robo-messages: reports, reminders, etc. These are really valuable to a few people, but add a lot more to wade through for everyone else trying to keep up.
There are certainly others, as well.
In short… there probably shouldn’t just be one thing. But, the cross-posting problem makes it hard to split up as a mailing list.
Enter Discourse ---------------
If you’ve talked with me about anything related to any of this in the past ten years, you probably already know that I like Discourse. It’s good software made by cool people. And, it’s entirely open source, with a SaaS business model but with real, usable releases. (No open core, no “open source theoretically but good luck”.) And, we have it in production in Fedora already, at https://discussion.fedoraproject.org, with categories for announcements, user help, project discussion, and social conversation — as well as special categories for dedicated workflows.
In Project Discussion, each different Fedora team can have its own tag, and you can subscribe to those that you’re interested in. Cross-posting is easy: tag a post with multiple teams.
Topics can be renamed, if needed, or split out into side-conversations. The long tangents from these conversations can actually be interesting on their own without distracting from the main topic. Moderation tools allow us to handle posts that are outside of expected Fedora contributor behavior, with varying levels of action as appropriate.
You can use markdown formatting, including images (with easy addition of alt text for people for whom images are a barrier). You can edit your posts to fix typos or correct mistakes. There are polls and lots of other useful features.
And, you can interact with it all by email. That interface isn’t perfect, but it’s way better than any other forum software I’ve seen. (I’ve written a guide: https://discussion.fedoraproject.org/t/25960) At the same time, your individual email address is hidden, so it won’t be a spam magnet (or a target for off-list harassment, a problem we unfortunately have with devel list).
That said, it is web-first software. (Or mobile, if that’s your thing.) That requires some adjustment, I know. I hope opening up a Fedora Discussion tab – or keeping one open — becomes an easy habit.
Not just Fedora ---------------
There’s a big trend towards Discourse in open source projects overall. Python and Gnome have both migrated entirely from their mailing lists. Ansible is working on it. Plus, there’s Rust, Kubernetes, Nextcloud, Flathub, Grafana, Home Assistant, KDE, and I’m sure many others.
Concrete proposal -----------------
I’m not suggesting we shut down devel list next week. And I think we’ll have some mailing lists for quite a long time. But, I think it’s time to start moving some specific things, with the eventual goal of closing every mailing list we can.
First, I’d like to move the Changes discussion. They will still be posted to devel-announce, but responses directed to Project Discussion in a new #changes tag. Ben tells me that this is a FESCo decision, which seems reasonable.
Second, I think other FESCo-related conversations should move. I hope this will reduce the urge to have back-and-forth exchanges in the tickets. For the Fedora Council, I set up a bot which automatically creates a discussion topic when a ticket is filed, leaving the ticket just for votes and recording of outcome. FESCo could use something similar.
Third, I’ll add a tag for general Fedora OS development conversation. Maybe “#devel”, but if someone has more narrow suggestions, I’ll take them.
Fourth, I’d like to update our documentation, process, and expectations for newcomers — say hello on Discussion (and Fedora Chat, if you like) rather than a mailing list. (I’d like to close the Fedora Join list at this point.)
Fifth, all packaging-related discussions (including the separate packaging mailing list). We already have a #package-maintainers tag with some existing discussion.
Sixth, automated posts, as much as we can. These should go to dedicated Workflow categories, where people who want can watch them but where they won’t overwhelm human interactions. People who want can watch them, and it’s easy to quote-reply into a new linked topic in the Project Discussions category.
And finally… shut down the devel list itself. Perhaps at the end of 2023?
We should also shut down all of the little lists that haven’t had anything but spam in the last two years. We could maybe do that sooner. We should stop creating new lists now — we can create new Discussion tags instead.
I expect the announcement lists to stay for the foreseeable future (although we might feed them from Discourse rather than the other way around). Other lists which are patches, commit messages, and other automations might stick around for a while — but really might be better served by a log aggregation and analysis system or something else.
Other teams who want to keep mailing lists can, but I’d like to move those too, and eventually I think we’ll want to shut them down too — or perhaps convert them to announcement lists.
Next steps ----------
I know this is a big change. I’ve been thinking of writing this message for a long time. I’d really like to convince everyone that it’s the right thing — or at least, an acceptable one.
What about specific decisions related to my proposal? For each:
Because altering the Changes process is a FESCo decision, I’ll file a ticket about that shortly.
FESCo moving their own other conversations is, of course, also a FESCo decision.
Assuming the first moves forward, I will create the general #devel tag (or other name we come up with) when I create the #change-proposal tag.
Moving the packaging list is a Packaging Committee decision.
Automated posts can be moved at any time. I can work with the people who own the generation of those reports to figure out a good answer for each.
The outcome for other team lists is up to each team.
And, I think shutting down devel overall is ultimately a Fedora Council decision.
For right now, though: let’s discuss — on the list!
On Thu, 20 Apr 2023 at 17:21, Matthew Miller mattdm@fedoraproject.org wrote:
It’s time to transform the Fedora devel list into something new
I hate to ask this but could you give a more summarized version of this email? I realize you had a lot of reasoning you wanted to cover on the why's but I frankly got lost several times. That makes it really hard not to respond in ways which are overly emotional and not helpful. Anything I wrote would start with me trying to summarize what was written but failing to do so, or I would end up trying to pick apart different paragraphs in non-helpful ways.
On Thu, Apr 20, 2023 at 05:39:51PM -0400, Stephen Smoogen wrote:
I hate to ask this but could you give a more summarized version of this email? I realize you had a lot of reasoning you wanted to cover on the why's but I frankly got lost several times. That makes it really hard not to respond in ways which are overly emotional and not helpful. Anything I wrote would start with me trying to summarize what was written but failing to do so, or I would end up trying to pick apart different paragraphs in non-helpful ways.
Sure. I realize it is quite long.
I am proposing that over the course of 2023, starting with the Changes process, we move Fedora development conversations from this mailing list to the Discourse-based Fedora Discussion.
Many Fedora folks, new and old, can't keep with this list. The number of participants is down over time (even as the number of threads has risen). Many teams are moving away from devel list anyway -- using various scattered bug trackers as their effective "forum".
Discourse gives us better tools for the conversations we need to have as a project. I know it takes some getting used to, but I strongly believe it will be worth it.
Devel list actually covers a lot of different topics. Discourse lets us categorize those better while still keeping it all together.
The first thing I suggest moving is discussion around proposed Changes. This is a FESCo decision. The rest I won't duplicate here.
On Thu, 20 Apr 2023 at 18:47, Matthew Miller mattdm@fedoraproject.org wrote:
On Thu, Apr 20, 2023 at 05:39:51PM -0400, Stephen Smoogen wrote:
I hate to ask this but could you give a more summarized version of this email? I realize you had a lot of reasoning you wanted to cover on the why's but I frankly got lost several times. That makes it really hard not to respond in ways which are overly emotional and not helpful. Anything I wrote would start with me trying to summarize what was written but
failing
to do so, or I would end up trying to pick apart different paragraphs in non-helpful ways.
Sure. I realize it is quite long.
I am proposing that over the course of 2023, starting with the Changes process, we move Fedora development conversations from this mailing list to the Discourse-based Fedora Discussion.
Many Fedora folks, new and old, can't keep with this list. The number of participants is down over time (even as the number of threads has risen). Many teams are moving away from devel list anyway -- using various scattered bug trackers as their effective "forum".
Discourse gives us better tools for the conversations we need to have as a project. I know it takes some getting used to, but I strongly believe it will be worth it.
Devel list actually covers a lot of different topics. Discourse lets us categorize those better while still keeping it all together.
The first thing I suggest moving is discussion around proposed Changes. This is a FESCo decision. The rest I won't duplicate here.
Thank you. I have a better understanding of where you are coming from, and what this meant to do. I don't like the solution, but I know all too well that the current mailman3 solution works on a wing and a prayer. It has been running an EOL version of the software for a long time and there are not enough infrastructure resources to do all the things that are needed for an upgrade AND keep builds going. I also understand that the general community of the lists has shrunk over the last 10 years with it becoming more and more 'the same old people complaining about the same old things'.
That said, I don't think I will be greatly active after the move. I have tried Discourse for a year, but have found it to be like every forum and BBS I have tried for the last 30 years.. frustrating and needy. I get tired and angry after 30 minutes and my replies start becoming the problem you don't want. [I realize this is how many people feel about email which causes them to drop out there.] If that lack of engagement requires me to orphan packages or other items, I completely understand.
On to, 20 huhti 2023, Stephen Smoogen wrote:
On Thu, 20 Apr 2023 at 18:47, Matthew Miller mattdm@fedoraproject.org wrote:
On Thu, Apr 20, 2023 at 05:39:51PM -0400, Stephen Smoogen wrote:
I hate to ask this but could you give a more summarized version of this email? I realize you had a lot of reasoning you wanted to cover on the why's but I frankly got lost several times. That makes it really hard not to respond in ways which are overly emotional and not helpful. Anything I wrote would start with me trying to summarize what was written but
failing
to do so, or I would end up trying to pick apart different paragraphs in non-helpful ways.
Sure. I realize it is quite long.
I am proposing that over the course of 2023, starting with the Changes process, we move Fedora development conversations from this mailing list to the Discourse-based Fedora Discussion.
Many Fedora folks, new and old, can't keep with this list. The number of participants is down over time (even as the number of threads has risen). Many teams are moving away from devel list anyway -- using various scattered bug trackers as their effective "forum".
Discourse gives us better tools for the conversations we need to have as a project. I know it takes some getting used to, but I strongly believe it will be worth it.
Devel list actually covers a lot of different topics. Discourse lets us categorize those better while still keeping it all together.
The first thing I suggest moving is discussion around proposed Changes. This is a FESCo decision. The rest I won't duplicate here.
Thank you. I have a better understanding of where you are coming from, and what this meant to do. I don't like the solution, but I know all too well that the current mailman3 solution works on a wing and a prayer. It has been running an EOL version of the software for a long time and there are not enough infrastructure resources to do all the things that are needed for an upgrade AND keep builds going. I also understand that the general community of the lists has shrunk over the last 10 years with it becoming more and more 'the same old people complaining about the same old things'.
That said, I don't think I will be greatly active after the move. I have tried Discourse for a year, but have found it to be like every forum and BBS I have tried for the last 30 years.. frustrating and needy. I get tired and angry after 30 minutes and my replies start becoming the problem you don't want. [I realize this is how many people feel about email which causes them to drop out there.] If that lack of engagement requires me to orphan packages or other items, I completely understand.
My main trouble with Discourse and other places where I try to help people with answers to their questions is that forums promote a drive-by questions without further engagement. This experience is opposite to what forum proponents are claiming but I see it pretty consistently on Discourse, on Stackoverflow sites, on Reddit and in many other places.
In my area, identity management and authentication, the topics are complex enough to want to help others but lack of further engagement simply kills any interest to use a particular discussion board. If people asking questions aren't interested in getting the answers or even tying in the ends for their own questions, it comes hard to keep an interest in helping those people again and again.
I can point you to one specific topic on Fedora Discourse as an example: https://discussion.fedoraproject.org/t/fedora-login-bug-having-a-128-charact...
I would have supposed that someone would follow-up, right? As a FreeIPA maintainer in Fedora, as an upstream FreeIPA contributor and a contact for security issues, I have never been contacted with either details for what the thread claims to happen or never got any follow-up on the thread to my comments.
This is an experience I want to avoid. If this is what Matthew is proposing a Fedora development discussions to be, then sorry, this is not an improvement.
On Fri, Apr 21, 2023 at 09:38:26AM +0300, Alexander Bokovoy wrote:
My main trouble with Discourse and other places where I try to help people with answers to their questions is that forums promote a drive-by questions without further engagement. This experience is opposite to what forum proponents are claiming but I see it pretty consistently on Discourse, on Stackoverflow sites, on Reddit and in many other places.
I think some of that is natural in any support forum. The same thing happens on the Fedora Users' mailing list.
In my area, identity management and authentication, the topics are complex enough to want to help others but lack of further engagement simply kills any interest to use a particular discussion board. If people asking questions aren't interested in getting the answers or even tying in the ends for their own questions, it comes hard to keep an interest in helping those people again and again.
I can point you to one specific topic on Fedora Discourse as an example: https://discussion.fedoraproject.org/t/fedora-login-bug-having-a-128-charact...
Well, in that case, I think the person was primarily interested in getting access to their account back, not the underlying bug. Justin helped them find where to file a ticket, and presumably everything was resolved from their point of view.
I would have supposed that someone would follow-up, right? As a FreeIPA maintainer in Fedora, as an upstream FreeIPA contributor and a contact for security issues, I have never been contacted with either details for what the thread claims to happen or never got any follow-up on the thread to my comments.
I really don't think that's a _tooling_ issue.
This is an experience I want to avoid. If this is what Matthew is proposing a Fedora development discussions to be, then sorry, this is not an improvement.
Well, no. It is not what I am proposing.
Am 21.04.23 um 01:34 schrieb Stephen Smoogen:
That said, I don't think I will be greatly active after the move. I have tried Discourse for a year, but have found it to be like every forum and BBS I have tried for the last 30 years.. frustrating and needy.
ACK.
- you do not see, what new answeres have arrived - you need to open and login into it, everytime your browser gets started - you need to actively look into any open TAB for any ML with theire own discourse-like-websolution - you are required to work with tools that specific service offers - "offline" breaks the workflow - you need a shitload of traffic to archive the same result, as you can have with one single email. - it takes longer to find something aso. aso.
and best of all:
- you (multiply this with any participant) need way more cpu power (at home and on the server) to archive the same results => Which is bad for the climate!!!
Mails get concentrated in one place in an efficient way, with a short and easy workflow in one system, we all need to work on a day-by-day basis, or do you know anyone who can uninstall his/her mailapp after the Fedora ML has moved to Discourse? I don't think so.
If "you",reader, have problems following the threads with your mailapp, which was the main argument here, get a better mailapp to handle it.
Best regards, Marius Schwarz
Once upon a time, Marius Schwarz fedoradev@cloud-foo.de said:
- you are required to work with tools that specific service offers
I think this is my biggest complaint with any web forum - unlike email, where users can choose clients that work the way they like, learning and customizing them, web forums force a use/workflow method based on what some developer likes. They also tend to be "locked in" to a singular UI, because any significant change after a critical mass of users is reached is too upsetting.
Also: with web forums, I feel my content is not my own. I can't automatically keep copies and archive them. For example, I can in seconds tell you when my first post here was (January 2010 in the current incarnation, September 2003 on the @redhat.com version), or even my first post to any @redhat.com list (wow, June 1996!).
Am 21.04.2023 um 14:27 schrieb Chris Adams linux@cmadams.net:
I can't automatically keep copies and archive them. For example, I can in seconds tell you when my first post here was
This is an extremely important criterion to consider when renewing our communications tools.
I can arrange information that I find important in a way that makes it quick and easy to find.
In discouse, this could perhaps be simulated by giving everyone a virtual repository for copies (links).
Once upon a time, Matthew Miller mattdm@fedoraproject.org said:
I am proposing that over the course of 2023, starting with the Changes process, we move Fedora development conversations from this mailing list to the Discourse-based Fedora Discussion.
I feel this is a case of trading one group of people (email list users) for a different group of people (web forum users). I have seen this done multiple times over the years, tried to follow a few times, and always dropped off fairly rapidly. I'm solidly in the "email list users" group.
Web forums far from fit how I use communication tools. For example, I'm highly keyboard driven and dislike lots of things that force me to use a mouse. I use mutt to read email, which works great. I have email filters to organize and save things, I can flag messages for later review/reference, and more.
Web forums also seem in my experience to be much more casual interaction, where people come and go for long stretches of time, much more than email lists. If I just don't click on your bookmark for a day or two, it's out of mind quickly and I might not come back for weeks or months (or ever, which is usually the case). And then even if I do come back, now I'm pretty much too far behind to ever catch up.
I guess you're hoping for enough overlap between email and web forum users to keep an active developer community... I wish you luck with that.
On 4/21/23 04:24, Chris Adams wrote:
Once upon a time, Matthew Miller mattdm@fedoraproject.org said:
I am proposing that over the course of 2023, starting with the Changes process, we move Fedora development conversations from this mailing list to the Discourse-based Fedora Discussion.
I feel this is a case of trading one group of people (email list users) for a different group of people (web forum users). I have seen this done multiple times over the years, tried to follow a few times, and always dropped off fairly rapidly. I'm solidly in the "email list users" group.
Discourse is very nice. It is open source, uses a reasonable license [0] (though [1] would be better), and is great for casual interactions that maybe spread out over time. Would highly recommend it for moderate size community discussion.
However, it doesn't seem like we can hack on it to better suite community needs, for example to have the same functionality as mailing lists[2]. It is not standards driven and is primarily developed by one company - something that follows Apache way[3] or has a community governance process would be better in the long term for a large project with many contributors who have technical expertise.
Email clients offer significant customizability that a one size fits all web interface cannot provide. Mailing list mode for Discourse is helpful, but not at the same level as email lists, where once one has gained sufficient knowledge, interaction can be done from the comfort of the client of your choice. As such, simply adopting it because it can be deployed may leave out many contributors, in particular those who drive development forward. Mailing lists are not perfect, but it is not clear Discourse is a good replacement for the devel list.
0) https://github.com/discourse/discourse/blob/main/LICENSE.txt 1) https://www.gnu.org/licenses/agpl-3.0.en.html 2) https://discourse.cmake.org/t/cmake-discourse-mailing-list-mode-incorrectly-... 3) https://apache.org/theapacheway/
On Fri, Apr 21, 2023 at 08:24:14AM +0300, Benson Muite wrote:
As such, simply adopting it because it can be deployed may leave out many contributors, in particular those who drive development forward.
I have made this point several times in other contexts; a new tool/workflow has to yield tangible improvements for the _existing_ contributor base. Otherwise you're just going to trade away _current_ contributors for the _possibility_ of attracting new ones.
(In a volunteer context, that is)
- Solomon
On Fri, Apr 21, 2023 at 08:24:14AM +0300, Benson Muite wrote:
However, it doesn't seem like we can hack on it to better suite community needs, for example to have the same functionality as mailing lists[2]. It is not standards driven and is primarily developed by one company - something that follows Apache way[3] or has a community governance process would be better in the long term for a large project with many contributors who have technical expertise.
We definitely _can_ hack on it to better fit community needs. Changes might not automatically get accepted, but we've got a good relationship and I don't expect any kind of antagonism if we have something important.
On 2) https://discourse.cmake.org/t/cmake-discourse-mailing-list-mode-incorrectly-... in particular... that's just the Cmake forum admin saying that the particular thing doesn't exist, not a Discourse dev saying they won't take a change.
Although on that specific change.... Discourse attaches List-Id and other standard email headers (as well as some specific X-Discourse headers) to each message. Changing the To: line to be some list address could be done with a plugin, but might actually have negative consequences for reliable delivery.
Email clients offer significant customizability that a one size fits all web interface cannot provide. Mailing list mode for Discourse is helpful, but not at the same level as email lists, where once one has
"Mailing list mode" was a specific thing in earlier versions of Discourse — it sent a notification for every message posted. This is kind of like going to Hyperkitty and saying "subscribe me to all 600 lists". I don't recommend that. Instead, choose specific tags that you want to subscribe to, just as you would subscribe to individual mailing lists.
I have a post about this and Fedora Discussion specifically:
https://discussion.fedoraproject.org/t/navigating-fedora-discussion-tags-cat...
On 21/04/2023 18:52, Matthew Miller wrote:
"Mailing list mode" was a specific thing in earlier versions of Discourse — it sent a notification for every message posted. This is kind of like going to Hyperkitty and saying "subscribe me to all 600 lists". I don't recommend that. Instead, choose specific tags that you want to subscribe to, just as you would subscribe to individual mailing lists.
It's still a thing in current versions, it just doesn't seem to be available in the Fedora installation.
It also doesn't have to send you everything as you can mute those things you don't want to include.
Having to positively opt in to certain tags seems like a terrible idea as you're bound to miss lots of things when people create new tags that you don't even know exist. I'd much rather get everything by default and then opt out of the things I'm not interested in.
Tom
On Fri, Apr 21, 2023 at 07:05:05PM +0100, Tom Hughes via devel wrote:
"Mailing list mode" was a specific thing in earlier versions of Discourse — it sent a notification for every message posted. This is kind of like going
It's still a thing in current versions, it just doesn't seem to be available in the Fedora installation.
It's hidden from the settings UI by default in new installs. Looks like it is still there as an option to enable, but I really do think it's kind of a trap.
If the category watch thing doesn't work for you (and others) for some reason, I can re-consider. If so, I might rename it to "email me everything that isn't muted" instead of "mailing list mode", though....
It also doesn't have to send you everything as you can mute those things you don't want to include.
That will at least keep you from getting thousands of Copr posts. :)
Having to positively opt in to certain tags seems like a terrible idea as you're bound to miss lots of things when people create new tags that you don't even know exist. I'd much rather get everything by default and then opt out of the things I'm not interested in.
In that case, subscribing by category should do the trick. Categories are not light-weight, so we're unlikely to create new ones without fanfare. It's not just one checkbox, but there are only a handful of main ones anyway.
On 4/21/23 20:52, Matthew Miller wrote:
On Fri, Apr 21, 2023 at 08:24:14AM +0300, Benson Muite wrote:
However, it doesn't seem like we can hack on it to better suite community needs, for example to have the same functionality as mailing lists[2]. It is not standards driven and is primarily developed by one company - something that follows Apache way[3] or has a community governance process would be better in the long term for a large project with many contributors who have technical expertise.
We definitely _can_ hack on it to better fit community needs. Changes might not automatically get accepted, but we've got a good relationship and I don't expect any kind of antagonism if we have something important.
Good working relationship is ok for small projects. Discourse has grown and adapted because of this. Discourse is fantastic as a forum which is not read on a daily basis but periodically as the need arises. Need stability and a governance model for critical infrastructure. Web first philosophy kills productivity that a primarily text driven workflow has for many people.
On 2) https://discourse.cmake.org/t/cmake-discourse-mailing-list-mode-incorrectly-... in particular... that's just the Cmake forum admin saying that the particular thing doesn't exist, not a Discourse dev saying they won't take a change.
Although on that specific change.... Discourse attaches List-Id and other standard email headers (as well as some specific X-Discourse headers) to each message. Changing the To: line to be some list address could be done with a plugin, but might actually have negative consequences for reliable delivery.
Email clients offer significant customizability that a one size fits all web interface cannot provide. Mailing list mode for Discourse is helpful, but not at the same level as email lists, where once one has
"Mailing list mode" was a specific thing in earlier versions of Discourse — it sent a notification for every message posted. This is kind of like going to Hyperkitty and saying "subscribe me to all 600 lists". I don't recommend that. Instead, choose specific tags that you want to subscribe to, just as you would subscribe to individual mailing lists.
I have a post about this and Fedora Discussion specifically:
https://discussion.fedoraproject.org/t/navigating-fedora-discussion-tags-cat...
This is helpful. Wish it were a magazine article. Those get read.
On Sat, Apr 22, 2023 at 07:01:06AM +0300, Benson Muite wrote:
https://discussion.fedoraproject.org/t/navigating-fedora-discussion-tags-cat...
This is helpful. Wish it were a magazine article. Those get read.
Interesting idea! I'll check with the Magazine team to see if they think it would be a good fit.
On Fri, Apr 21, 2023 at 12:24 AM Benson Muite benson_muite@emailplus.org wrote:
On 4/21/23 04:24, Chris Adams wrote:
Once upon a time, Matthew Miller mattdm@fedoraproject.org said:
I am proposing that over the course of 2023, starting with the Changes process, we move Fedora development conversations from this mailing list to the Discourse-based Fedora Discussion.
I feel this is a case of trading one group of people (email list users) for a different group of people (web forum users). I have seen this done multiple times over the years, tried to follow a few times, and always dropped off fairly rapidly. I'm solidly in the "email list users" group.
Discourse is very nice. It is open source, uses a reasonable license [0] (though [1] would be better), and is great for casual interactions that maybe spread out over time. Would highly recommend it for moderate size community discussion.
However, it doesn't seem like we can hack on it to better suite community needs, for example to have the same functionality as mailing lists[2]. It is not standards driven and is primarily developed by one company - something that follows Apache way[3] or has a community governance process would be better in the long term for a large project with many contributors who have technical expertise.
Email clients offer significant customizability that a one size fits all web interface cannot provide. Mailing list mode for Discourse is helpful, but not at the same level as email lists, where once one has gained sufficient knowledge, interaction can be done from the comfort of the client of your choice. As such, simply adopting it because it can be deployed may leave out many contributors, in particular those who drive development forward. Mailing lists are not perfect, but it is not clear Discourse is a good replacement for the devel list.
As Matthew stated, Ben has measured it and fewer people are participating on the mailing list over time. We are already leaving out many contributors. Those conversations are largely moving to issue trackers, which are also not perfect but are clearly more appealing than email for many people. Discourse has the potential to be a more attractive alternative than both email and issue trackers. To me this seems like a solid strategy for reversing the trend and getting more people participating in development discussions.
- https://github.com/discourse/discourse/blob/main/LICENSE.txt
- https://www.gnu.org/licenses/agpl-3.0.en.html
https://discourse.cmake.org/t/cmake-discourse-mailing-list-mode-incorrectly-... 3) https://apache.org/theapacheway/ _______________________________________________ devel mailing list -- devel@lists.fedoraproject.org To unsubscribe send an email to devel-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
On Fri, 2023-04-21 at 15:39 -0500, Carl George wrote:
As Matthew stated, Ben has measured it and fewer people are participating on the mailing list over time. We are already leaving out many contributors.
This is an interpretation, but are we sure we are missing them because the mailing list is uninviting, and not just because the pool of interested people is shrinking?
Those conversations are largely moving to issue trackers, which are also not perfect but are clearly more appealing than email for many people.
Issue trackers are a pretty good way to deal with issues, we should have a better issue tracker that makes those conversations better, not discourage them to make everything flow in non-descript mailing lists or forums...
Discourse has the potential to be a more attractive alternative than both email and issue trackers.
And less attractive to people that work better with mailing lists and issue trackers...
To me this seems like a solid strategy for reversing the trend and getting more people participating in development discussions.
I really dislike this fixation on numbers. We need higher quality and we need to discuss what is really needed. Numbers shouldn't be priority number one, unless there are other underlying issues.
Simo.
On Thu, Apr 20, 2023 at 08:24:48PM -0500, Chris Adams wrote:
Once upon a time, Matthew Miller mattdm@fedoraproject.org said:
I am proposing that over the course of 2023, starting with the Changes process, we move Fedora development conversations from this mailing list to the Discourse-based Fedora Discussion.
I feel this is a case of trading one group of people (email list users) for a different group of people (web forum users).
I don't think this is _really_ a "there are two kinds of people in the world..." situation. Of course there are some people who have preferences (strong or weak) for one or the other, and completely legitimate pros and cons to each. But I don't want to "trade" anyone. I'd like to bring everyone along.
I have seen this done multiple times over the years, tried to follow a few times, and always dropped off fairly rapidly. I'm solidly in the "email list users" group.
Is there anything which could be different this time which would make it better for you?
On Fri, 2023-04-21 at 10:44 -0400, Matthew Miller wrote:
On Thu, Apr 20, 2023 at 08:24:48PM -0500, Chris Adams wrote:
Once upon a time, Matthew Miller mattdm@fedoraproject.org said:
I am proposing that over the course of 2023, starting with the Changes process, we move Fedora development conversations from this mailing list to the Discourse-based Fedora Discussion.
I feel this is a case of trading one group of people (email list users) for a different group of people (web forum users).
I don't think this is _really_ a "there are two kinds of people in the world..." situation. Of course there are some people who have preferences (strong or weak) for one or the other, and completely legitimate pros and cons to each. But I don't want to "trade" anyone. I'd like to bring everyone along.
I have seen this done multiple times over the years, tried to follow a few times, and always dropped off fairly rapidly. I'm solidly in the "email list users" group.
Is there anything which could be different this time which would make it better for you?
So I am trying to see how it works to follow those discussions via email (sorry this is the only way, following via web simply doesn't work for me).
So I registered the account, added the email I want to get notifications at, and selected a few topics.
First impressions.
It is absolutely confusing to figure out how to watch topics. If you select a category and a topic you do not get the notification bell to watch them. If you select just a topic you get it. Topics are all in random order. When you select some topic by searching you sometimes are then proposed a different one (? renames ?)
I found no way to watch all, and let my client sort it out ... which would need client filtering, because the stupid gmail filtering can't handle header fields (#@#%$@#). And if you can't watch all it means you can watch only know topics, and anything new will simply be missed.
It would be useful to be able to say "watch all except these specific topics", which are the ones you realize are uninteresting to you and explicitly start to filter out. I could do that filtering after receiving though, so being able to watch all is the minimum requirement to be able to follow anything.
[late edit: in a plot twist, if you select a category, then you get a new meta-tag call all-tags, and you can watch that ... I selected that now and will see if I can block specific tags somehow later or filter them on my end]
I couldn't find a place to list all the topics I am watching ... only way I found so far is to click on all the tags in a notification to see if any of them has the bell thingy ...
On the actual notifications I am receiving:
They come several minutes (at least 5 minutes, as the email is *sent* that much later, and the sent date is set to when the email is sent, not to when the post is made) after the actual message is posted in discourse, I do not care much, I generally read asynchronously as well, but it is sometimes annoying not to be able to establish the real time a post was made.
The only way to know who posted is by looking at the From field, where the description of the notification email address is changed to include the display name of the poster. That is a bit confusing at times.
There is no formatting in the mail that tells me who is someone replying to, or which message in the thread it was being replied to (I disabled sending me the whole thread with each notification, I may re- enable it).
The test part so far is otherwise decently rendered, and for image posts it is clear enough that there is something to look at in the html part. *however* the images are not embedded in the email, so all that information is unavailable offline or for archival (and in my configuration requires to actively pull images as I configured my client to not pull 3rd party content automatically for privacy and security reasons).
I have not tried to reply to anything, so I do not know how that will fare.
Il 21/04/23 17:37, Simo Sorce ha scritto:
On Fri, 2023-04-21 at 10:44 -0400, Matthew Miller wrote:
On Thu, Apr 20, 2023 at 08:24:48PM -0500, Chris Adams wrote:
Once upon a time, Matthew Miller mattdm@fedoraproject.org said:
I am proposing that over the course of 2023, starting with the Changes process, we move Fedora development conversations from this mailing list to the Discourse-based Fedora Discussion.
I feel this is a case of trading one group of people (email list users) for a different group of people (web forum users).
I don't think this is _really_ a "there are two kinds of people in the world..." situation. Of course there are some people who have preferences (strong or weak) for one or the other, and completely legitimate pros and cons to each. But I don't want to "trade" anyone. I'd like to bring everyone along.
I have seen this done multiple times over the years, tried to follow a few times, and always dropped off fairly rapidly. I'm solidly in the "email list users" group.
Is there anything which could be different this time which would make it better for you?
So I am trying to see how it works to follow those discussions via email (sorry this is the only way, following via web simply doesn't work for me).
So I registered the account, added the email I want to get notifications at, and selected a few topics.
First impressions.
...
Discourse docs say that it can be fully used like a mailing list, but it mention a "mailing list" mode that I can't find in Fedora Discourse profile preferences... is that feature available?
Mattia
On Fri, Apr 21, 2023 at 11:37:20AM -0400, Simo Sorce wrote:
So I registered the account, added the email I want to get notifications at, and selected a few topics.
First impressions.
It is absolutely confusing to figure out how to watch topics. If you select a category and a topic you do not get the notification bell to watch them. If you select just a topic you get it. Topics are all in random order. When you select some topic by searching you sometimes are then proposed a different one (? renames ?)
A terminology thing: I think what you are calling "topics" discourse calls "tags". Each discussion or thread is what discourse calls a "topic".
Watches are by category, by tag, or by individual topic.
I found no way to watch all, and let my client sort it out ... which would need client filtering, because the stupid gmail filtering can't handle header fields (#@#%$@#).
Watching at the category level is probably the closest thing here. That is, watch Project Discussion and News & Announcements at least. That will include all topics under those categories regardless of tag.
As for Gmail, see https://gist.github.com/tpopela/e2f17bf8eac15bee734b993e170f4dfa. I'm trying to get Tomas to write a Discourse post about that.
They come several minutes (at least 5 minutes, as the email is *sent* that much later, and the sent date is set to when the email is sent, not to when the post is made) after the actual message is posted in discourse, I do not care much, I generally read asynchronously as well, but it is sometimes annoying not to be able to establish the real time a post was made.
Oh that's interesting. I'll bring that up. The five minute delay is actually a site-wide configuration option: it gives time for the poster to make any quick typo fixes, add tags they forgot, etc. before the mail is sent. (The default is 10.)
The only way to know who posted is by looking at the From field, where the description of the notification email address is changed to include the display name of the poster. That is a bit confusing at times.
There is no formatting in the mail that tells me who is someone replying to, or which message in the thread it was being replied to (I disabled sending me the whole thread with each notification, I may re- enable it).
Do you have "Include an excerpt of replied to post in emails" checked? Does that help?
The test part so far is otherwise decently rendered, and for image posts it is clear enough that there is something to look at in the html part. *however* the images are not embedded in the email, so all that information is unavailable offline or for archival (and in my configuration requires to actively pull images as I configured my client to not pull 3rd party content automatically for privacy and security reasons).
Reasonably enough. There might be an option to embed images -- I'll look. For what it's worth, the images should all (and only) come from the dedicated CDN site for our hosting, and there's no linked tracking on our side or Discourse's. There's probably logs somewhere, though.
On Fri, 2023-04-21 at 14:27 -0400, Matthew Miller wrote:
On Fri, Apr 21, 2023 at 11:37:20AM -0400, Simo Sorce wrote:
So I registered the account, added the email I want to get notifications at, and selected a few topics.
First impressions.
It is absolutely confusing to figure out how to watch topics. If you select a category and a topic you do not get the notification bell to watch them. If you select just a topic you get it. Topics are all in random order. When you select some topic by searching you sometimes are then proposed a different one (? renames ?)
A terminology thing: I think what you are calling "topics" discourse calls "tags". Each discussion or thread is what discourse calls a "topic".
Sorry, I did mean tags.
Watches are by category, by tag, or by individual topic.
I found no way to watch all, and let my client sort it out ... which would need client filtering, because the stupid gmail filtering can't handle header fields (#@#%$@#).
Watching at the category level is probably the closest thing here. That is, watch Project Discussion and News & Announcements at least. That will include all topics under those categories regardless of tag.
As for Gmail, see https://gist.github.com/tpopela/e2f17bf8eac15bee734b993e170f4dfa. I'm trying to get Tomas to write a Discourse post about that.
I've seen this before and it is a total monstrosity that works for those that use the Web UI, but for those that pulls from IMAP like me just makes everything worse, as those scripts run with delay, no thanks. I expect they will also break suddenly whenever gmail decides to change something, so I am not investing my time in that at all. Note that I use gmail only because I am forced to, so I have no desire in further exploring the matter, personally speaking, I have other tools to cope with filtering if I need to.
They come several minutes (at least 5 minutes, as the email is *sent* that much later, and the sent date is set to when the email is sent, not to when the post is made) after the actual message is posted in discourse, I do not care much, I generally read asynchronously as well, but it is sometimes annoying not to be able to establish the real time a post was made.
Oh that's interesting. I'll bring that up. The five minute delay is actually a site-wide configuration option: it gives time for the poster to make any quick typo fixes, add tags they forgot, etc. before the mail is sent. (The default is 10.)
The only way to know who posted is by looking at the From field, where the description of the notification email address is changed to include the display name of the poster. That is a bit confusing at times.
There is no formatting in the mail that tells me who is someone replying to, or which message in the thread it was being replied to (I disabled sending me the whole thread with each notification, I may re- enable it).
Do you have "Include an excerpt of replied to post in emails" checked? Does that help?
I have it, it seem it does nothing, as no excerpt is added.
The test part so far is otherwise decently rendered, and for image posts it is clear enough that there is something to look at in the html part. *however* the images are not embedded in the email, so all that information is unavailable offline or for archival (and in my configuration requires to actively pull images as I configured my client to not pull 3rd party content automatically for privacy and security reasons).
Reasonably enough. There might be an option to embed images -- I'll look. For what it's worth, the images should all (and only) come from the dedicated CDN site for our hosting, and there's no linked tracking on our side or Discourse's. There's probably logs somewhere, though.
The privacy statement is about why I do not configure my client to download in general, not referred to discourse specifically. Point is the image is not embedded, so you need to be online to be able to pull at all whenever you read the message. It also means if the CDN removes/renames that content at some point in the future, the message sitting in my mailbox is now broken.
-- Matthew Miller mattdm@fedoraproject.org Fedora Project Leader _______________________________________________ devel mailing list -- devel@lists.fedoraproject.org To unsubscribe send an email to devel-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
Once upon a time, Matthew Miller mattdm@fedoraproject.org said:
On Thu, Apr 20, 2023 at 08:24:48PM -0500, Chris Adams wrote:
I have seen this done multiple times over the years, tried to follow a few times, and always dropped off fairly rapidly. I'm solidly in the "email list users" group.
Is there anything which could be different this time which would make it better for you?
I really can't imagine a change for me (and I apologize if that sounds really "grumpy old man"... which I guess it starting to apply to me, since I was in college when a friend told me about some guy in Finland saying "hey Minix people..."). It really comes down to how I use a computer I guess; I am highly keyboard-focused, and I haven't seen a web forum yet that can handle that. Some have a few keyboard shortcuts, but they rarely fill the whole use and often are not well-maintained.
Email lets me have full control of how I consume it. I can sort it my way, save what I want and delete the rest, flag things for more review, etc. Web forums force me to consume their content their way, and then when I maybe have a way to deal with it, they change things. Also, I can easily edit email posts in vi until I get my message the way I want (for example, this paragraph started out as a sentence further down the message :) ).
So for web-based forums and such, they are very casual use for me, where I might drop in occasionally, but mostly just when searching for info. When I've tried web forums before (like when Red Hat killed off their mailing lists), I tend to lose interest and stop going pretty quickly (maybe it's an ADHD thing there, I don't know). I have a bookmark folder of a handful of web forums, and when I look at it, I mostly see sites I haven't visited in months or years.
The only GUI-based communication tools I have stuck with are Slack and Discord, which I can run largely with a keyboard. Even there, I'm only in a small number of servers (Slack is mostly just for work at this point).
I tried loading https://discussion.fedoraproject.org/ in Lynx, but it's way too "busy" to be able to visually browse it, and of course it has the "best viewed with JavaScript enabled" tag, which is usually kiss-of-death for Lynx users (in fact, I couldn't see a way to log in or post/comment).
I'm a small-time packager in Fedora, just a few leaf packages, but I do try to contribute to development discussions based on my experience and use cases. I don't like dropping out of it, but that just feels like the likely outcome.
On Fri, Apr 21, 2023 at 11:02:15AM -0500, Chris Adams wrote:
I really can't imagine a change for me (and I apologize if that sounds really "grumpy old man"... which I guess it starting to apply to me, since I was in college when a friend told me about some guy in Finland saying "hey Minix people..."). It really comes down to how I use a computer I guess; I am highly keyboard-focused, and I haven't seen a web forum yet that can handle that. Some have a few keyboard shortcuts, but they rarely fill the whole use and often are not well-maintained.
Accessiblity is important to Fedora, and I take this seriously. For Discourse, hit the ? key to bring up the page describing keyboard shortcuts.
If you find something you can't do, I'll report it as a priority bug.
Email lets me have full control of how I consume it. I can sort it my way, save what I want and delete the rest, flag things for more review, etc. Web forums force me to consume their content their way, and then when I maybe have a way to deal with it, they change things. Also, I can easily edit email posts in vi until I get my message the way I want (for example, this paragraph started out as a sentence further down the message :) ).
It's true that the composer is a web thing rather than your own editor (although there's browser plugins for that...). And it's also true that the software changes and not necessarily on your timeline. But with that in mind:
* Discourse does let you compose and save draft messages
* There's a handy "bookmark" feature which I use all the time. You can give a reason for bookmarking something so you remember later, and give it a future time to notify you.
I tried loading https://discussion.fedoraproject.org/ in Lynx, but it's way too "busy" to be able to visually browse it, and of course it has the "best viewed with JavaScript enabled" tag, which is usually kiss-of-death for Lynx users (in fact, I couldn't see a way to log in or post/comment).
I'm actually pretty impressed with how decent it is in elinks for a modern website. I do think you need Javascript to post, though. (And our elinks does not seem to be built with such support, which probably isn't sufficient anyway, alas.)
On 4/21/23 14:05, Matthew Miller wrote:
Accessiblity is important to Fedora, and I take this seriously. For Discourse, hit the ? key to bring up the page describing keyboard shortcuts.
One thing I don't care for when it comes to web apps and keyboard shortcuts is that they are non-standard. When I can process communications in my mail client, all mail uses the same keyboard shortcuts, no matter which site it came from. With web apps, every web app has it's own keyboard shortcuts, which makes learning them all difficult.
On Sun, Apr 23, 2023 at 06:11:45PM +0000, Randy Barlow via devel wrote:
On 4/21/23 14:05, Matthew Miller wrote:
Accessiblity is important to Fedora, and I take this seriously. For Discourse, hit the ? key to bring up the page describing keyboard shortcuts.
One thing I don't care for when it comes to web apps and keyboard shortcuts is that they are non-standard. When I can process communications in my mail client, all mail uses the same keyboard shortcuts, no matter which site it came from. With web apps, every web app has it's own keyboard shortcuts, which makes learning them all difficult.
Yeah, that can be frustrating. But in this case at least, at least it'll be the same keys across various Discourse sites, and as noted it's pretty clearly emerging as the most common tool for open source discussion forums.
I promise not to re-configure the keybindings for our site. :)
Matthew Miller writes:
I don't think this is _really_ a "there are two kinds of people in the world..." situation. Of course there are some people who have preferences (strong or weak) for one or the other, and completely legitimate pros and cons to each. But I don't want to "trade" anyone. I'd like to bring everyone along.
I reread the thread-starter. I did not get that impression. What I read was: we'd like to replace the mailing list with discourse. That's it.
There was some mention of discourse's mail integration, but it was clearly "mark the checkbox" type of thing.
I have seen this done multiple times over the years, tried to follow a few times, and always dropped off fairly rapidly. I'm solidly in the "email list users" group.
Is there anything which could be different this time which would make it better for you?
I do not believe that any web-based discussion forum would ever be comparable to a mailing list, The two environments are worlds apart, and irreconcilable, that's simply the way it is. They work in fundamentally different ways. I typed and deleted about two paragraphs' worth of my explanations why web-based forums are not a replacement for mailing list, I changed my mind and concluded that this kind of advocacy won't matter much. In the end, whatever happens, happens. The chips will fall where they may.
On Thu, Apr 20, 2023 at 8:25 PM Chris Adams linux@cmadams.net wrote:
Once upon a time, Matthew Miller mattdm@fedoraproject.org said:
I am proposing that over the course of 2023, starting with the Changes process, we move Fedora development conversations from this mailing list to the Discourse-based Fedora Discussion.
I feel this is a case of trading one group of people (email list users) for a different group of people (web forum users). I have seen this done multiple times over the years, tried to follow a few times, and always dropped off fairly rapidly. I'm solidly in the "email list users" group.
Web forums far from fit how I use communication tools. For example, I'm highly keyboard driven and dislike lots of things that force me to use a mouse. I use mutt to read email, which works great. I have email filters to organize and save things, I can flag messages for later review/reference, and more.
Web forums also seem in my experience to be much more casual interaction, where people come and go for long stretches of time, much more than email lists. If I just don't click on your bookmark for a day or two, it's out of mind quickly and I might not come back for weeks or months (or ever, which is usually the case). And then even if I do come back, now I'm pretty much too far behind to ever catch up.
That cuts both ways.
If I just don't click on [email filter folder] for a day or two, it's out of mind quickly and I might not come back for weeks or months. And then even if I do come back, now I'm pretty much too far behind to ever catch up.
I guess you're hoping for enough overlap between email and web forum users to keep an active developer community... I wish you luck with that.
-- Chris Adams linux@cmadams.net _______________________________________________ devel mailing list -- devel@lists.fedoraproject.org To unsubscribe send an email to devel-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
I dont mind discourse for a forum, although it's "one long list to rule them all" is frankly annoying at times and reminds me too much of the modern social media UX where you have to scroll and scroll to find something you care about. Sometimes, someone only cares about one subtopic. Yes I know this can be somewhat addressed with proper tagging, but that takes constant effort by everyone involved to make that useful. Most users wont use them, so its up to mods or other site users to constantly be back filling that information. It's a constant effort that must always be made to keep things orderly. It exchanges immediate convenience for recent information for a more long term effort to keep older information as easily accessible as it would have been in a mailing list or a classic forum/sub-forum style structure.
And speaking of older information... there's another problem with them in that aspect... the ability to work with them offline or local copy is not possible with a SAAS solution like discourse.
As I do a lot of historical research in open source and actively archiving what I can for future people. This is something I focus on and while I know there's not many of us that are doing it, it's still a thing for some of us. Working with old Distros and trying to research how we got from there to here has only been possible because people back in the day archived mailing lists and things like sunsite.unc.edu I can scrape a modern mailing lists for reference later, and pull up mailing lists that others have archived before me.
In an effort to be more efficient and "modern", are we taking away that possibility for the next generation? Using a SAAS solution doesn't seem to make that possible, but maybe I'm wrong and there is a way that I dont know about. Will there be an effort to export a PII sanitized database for people to use as an offline or local reference.
I'm not saying that we have to keep using the same tools we have in the past, if new tools can offer us new abilities that's great. But I'd appreciate it if there was a way to move forward with new tools while not taking away the abilities that older tools gave us to archive things for the future.
JT
On Thu, Apr 20, 2023 at 05:53:34PM -0400, JT wrote:
Sometimes, someone only cares about one subtopic. Yes I know this can be somewhat addressed with proper tagging, but that takes constant effort by everyone involved to make that useful. Most users wont use them, so its up to mods or other site users to constantly be back filling that information. It's a constant effort that must always be made to keep things orderly. It exchanges immediate convenience for recent information for a more long term effort to keep older information as easily accessible as it would have been in a mailing list or a classic forum/sub-forum style structure.
For tagging, in the Project Discussion category, each topic requires at least one tag, and all of the available tags are from a relatively-short list meant to correspond directly to active project teams. You can think of each of these as a kind of mailing list — you can subscribe to or mute each of these tags.
For example, you can find docs team topics at https://discussion.fedoraproject.org/tag/docs-team
In my experience in the last year or so with this structure, it works well and doesn't require a lot of maintenance.
The Ask Fedora category uses looser tagging generally based around topics, and c
For older information, Discourse has significantly better search than Hyperkitty provides.
And speaking of older information... there's another problem with them in that aspect... the ability to work with them offline or local copy is not possible with a SAAS solution like discourse.
That's basically true, although there is an API and it would be theoretically possible to make an offline client of some sort. We also have automatic backups so that all of the data is available in a Fedora-controlled way.
As I do a lot of historical research in open source and actively archiving what I can for future people. This is something I focus on and while I know there's not many of us that are doing it, it's still a thing for some of us. Working with old Distros and trying to research how we got from there to here has only been possible because people back in the day archived mailing lists and things like sunsite.unc.edu I can scrape a modern mailing lists for reference later, and pull up mailing lists that others have archived before me.
In an effort to be more efficient and "modern", are we taking away that possibility for the next generation?
I care about this too. I don't think we are taking away the possibility of archival research.
But, also: "efficient and modern" aren't in my reasons for suggesting this.
Using a SAAS solution doesn't seem to make that possible, but maybe I'm wrong and there is a way that I dont know about. Will there be an effort to export a PII sanitized database for people to use as an offline or local reference.
I don't have an effort like that planned, but I would not be opposed to someone who wants to work on that.
For example, you can find docs team topics at
https://discussion.fedoraproject.org/tag/docs-team
In my experience in the last year or so with this structure, it works
well and doesn't require a lot of maintenance.
Cool, a few of the discourse forums I'm on it's been a problem, but if our community is able to keep up with it then that's great to hear.
And speaking of older information... there's another problem with them
in that aspect... the ability to work with them offline or local copy is not possible with a SAAS solution like discourse.
That's basically true, although there is an API and it would be
theoretically possible to make an offline client of some sort. We also have automatic backups so that all of the data is available in a Fedora-controlled way.
Using a SAAS solution doesn't seem to make that possible, but maybe I'm wrong
and there is a way that I dont know about. Will there be an effort to export a PII sanitized database for people to use as an offline or local reference.
I don't have an effort like that planned, but I would not be opposed to
someone who wants to work on that.
I'm still getting settled in to a new job after a few months of being without work, and I need to build another safety net for myself for the next transitional period I face, but when I've gotten that taken care of, I'd be interested in funding some dev work towards making that a problem that we have a solution to.
JT
On Thu, Apr 20, 2023 at 7:00 PM Matthew Miller mattdm@fedoraproject.org wrote:
On Thu, Apr 20, 2023 at 05:53:34PM -0400, JT wrote:
Sometimes, someone only cares about one subtopic. Yes I know this can be somewhat addressed with proper tagging, but that takes constant effort by everyone involved to make that useful. Most users wont use them, so its
up
to mods or other site users to constantly be back filling that information. It's a constant effort that must always be made to keep things orderly. It exchanges immediate convenience for recent
information
for a more long term effort to keep older information as easily
accessible
as it would have been in a mailing list or a classic forum/sub-forum
style
structure.
For tagging, in the Project Discussion category, each topic requires at least one tag, and all of the available tags are from a relatively-short list meant to correspond directly to active project teams. You can think of each of these as a kind of mailing list — you can subscribe to or mute each of these tags.
For example, you can find docs team topics at https://discussion.fedoraproject.org/tag/docs-team
In my experience in the last year or so with this structure, it works well and doesn't require a lot of maintenance.
The Ask Fedora category uses looser tagging generally based around topics, and c
For older information, Discourse has significantly better search than Hyperkitty provides.
And speaking of older information... there's another problem with them in that aspect... the ability to work with them offline or local copy is not possible with a SAAS solution like discourse.
That's basically true, although there is an API and it would be theoretically possible to make an offline client of some sort. We also have automatic backups so that all of the data is available in a Fedora-controlled way.
As I do a lot of historical research in open source and actively
archiving
what I can for future people. This is something I focus on and while I know there's not many of us that are doing it, it's still a thing for
some
of us. Working with old Distros and trying to research how we got from there to here has only been possible because people back in the day archived mailing lists and things like sunsite.unc.edu I can scrape a modern mailing lists for reference later, and pull up mailing lists that others have archived before me.
In an effort to be more efficient and "modern", are we taking away that possibility for the next generation?
I care about this too. I don't think we are taking away the possibility of archival research.
But, also: "efficient and modern" aren't in my reasons for suggesting this.
Using a SAAS solution doesn't seem to make that possible, but maybe I'm wrong and there is a way that I dont know about. Will there be an effort to export a PII sanitized database for people to use as an offline or
local
reference.
I don't have an effort like that planned, but I would not be opposed to someone who wants to work on that.
-- Matthew Miller mattdm@fedoraproject.org Fedora Project Leader _______________________________________________ devel mailing list -- devel@lists.fedoraproject.org To unsubscribe send an email to devel-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
On Thu, 2023-04-20 at 18:59 -0400, Matthew Miller wrote:
On Thu, Apr 20, 2023 at 05:53:34PM -0400, JT wrote:
Sometimes, someone only cares about one subtopic. Yes I know this can be somewhat addressed with proper tagging, but that takes constant effort by everyone involved to make that useful. Most users wont use them, so its up to mods or other site users to constantly be back filling that information. It's a constant effort that must always be made to keep things orderly. It exchanges immediate convenience for recent information for a more long term effort to keep older information as easily accessible as it would have been in a mailing list or a classic forum/sub-forum style structure.
For tagging, in the Project Discussion category, each topic requires at least one tag, and all of the available tags are from a relatively-short list meant to correspond directly to active project teams. You can think of each of these as a kind of mailing list — you can subscribe to or mute each of these tags.
For example, you can find docs team topics at https://discussion.fedoraproject.org/tag/docs-team
In my experience in the last year or so with this structure, it works well and doesn't require a lot of maintenance.
Well, I just opened the tags box and typed 'qa', and got...#fedora-qa , #qa, and #qa-team . So it looks like some maintenance might be in order. :D Is there any way to 'guide' people to use 'standard' tags?
On Thu, Apr 20, 2023 at 06:01:51PM -0700, Adam Williamson wrote:
Well, I just opened the tags box and typed 'qa', and got...#fedora-qa , #qa, and #qa-team . So it looks like some maintenance might be in order. :D Is there any way to 'guide' people to use 'standard' tags?
Ah, see, this is a great example of a tangential subthread....
I'm still trying to figure out the best approach to one issue with tags since we merged in Ask Fedora.
Tags span across categories, including tag watches (that is, subscriptions). This means there's no way to subscribe to notifications for, e.g., #server in Project Discussion and _not_ to those in Ask. Ask can be kind of firehose of end-user queries, I want to make sure we have that option.
So, there is #server-wg (which we're probably going to rename to #server-devel to be less obtuse) for that. These "team" tags can only be used in the Project Discussion category (and Announcements). For qa, that's #qa-team. Then, there are some things that are tagged #fedora-qa in Ask.
"fedora-qa" is not particularly consistent, but #qa alone is kind of ambiguous and would get misused. As it was, #qa was set up as a synonym of #qa-team -- that's unnecessary and I've removed it.
Overall, tags in Ask are more of a wild territory. We discussed and have some general consensus on how they should be used: https://discussion.fedoraproject.org/t/how-should-we-use-tags-in-the-ask-fed... and I do a kind of lackadaisical periodic curation. (Tags can be created by anyone with a certain forum trust level, which scales with participation. But they can only be edited and removed by admins.)
(Although actually the emails triggered by notifications _do_ include `X-Discourse-Tags` and `X-Discourse-Category` headers, so if that's the your primary way of interacting, you _can_ do something with them. But it doesn't apply to on-site notifications.)
So, I've been using email and lists for... 25+ years now.
For me personally (and I think many of the other survivors you mention) we have carefully tuned filters and email clients and can read/reply to lists with ease and forums are new and anoying because you have to read them with the interface the forum has instead of your trusty mail client.
That said, for everyone except us, mail lists are much worse for all the reasons you list.
I'm willing to try to learn new tricks and move things to discourse.
kevin
On Thu, 2023-04-20 at 14:56 -0700, Kevin Fenzi wrote:
So, I've been using email and lists for... 25+ years now.
Same here…
For me personally (and I think many of the other survivors you mention) we have carefully tuned filters and email clients and can read/reply to lists with ease and forums are new and anoying because you have to read them with the interface the forum has instead of your trusty mail client.
In contrast, I've largely given up on staying on top of mailing lists for years now, so it’s not just an age thing (or maybe it is 🤔). Even with filters in place, the amount landing in say devel@ is just too overwhelming for me – a mailing list is a very coarse thing and I don’t find subject lines helpful to scan visually to trim down the deluge to manageable levels.
That said, for everyone except us, mail lists are much worse for all the reasons you list.
I'm willing to try to learn new tricks and move things to discourse.
I don’t know if using a forum software like Discourse could help, but I guess there’s something to be said about finer-grained “tags” or “topics” – if they are used properly and I don’t see us employing librarians to ensure that 😉. If a “change in venue” would let me stay on top things better, I would be very happy.
Nils
On Thu, Apr 20, 2023 at 4:21 PM Matthew Miller mattdm@fedoraproject.org wrote:
It’s time to transform the Fedora devel list into something new
My first post on this list was over 19 years ago. (It was about Bugzilla. I was a fan!) Ever since those early days, devel list has been the heart and center of Fedora activity. Now, I hope to convince all of us here that it’s time for something different.
As it is, devel list is too much for many people to follow — people we’d like to have around. It covers many different things at once, yet also drives us towards more scattered communications. Our infamous mega-threads are not really effective for getting to community consensus, and tend to bring out the worst in us.
I propose that we transition devel list, and eventually most of our mailing lists, to Fedora Discussion (our Discourse-powered forum).
I know this is a big change, but, hear me out…
I love this idea.
We’re missing people
A Mastodon post from long-time Fedora contributor Major Hayden got me thinking:
How do people make so much time available for mailing list discourse?
Once I ensure my team has the technical guidance they need and I work through the tasks of work that I owe other people, I take a look at the mailing list and say: "Oh my gosh, what the heck happened here?" Then the discussion goes further off the rails while I'm typing out a reply and my reply is no longer relevant.
I know many Fedora folks, old-school and new, for whom devel list is just too much. Some of it is the sheer volume, but this “off the rails” tendency is real — threads drift, get into back-and-forth debates about particular details, etc. And… some people aren’t here because — in contrast with our “Friends” foundation, it isn’t always a nice place to be (and mailing lists don’t provide many tools for moderation, except the big hammer of outright bans).
Ben Cotton recently did some basic analysis on devel list traffic over time, and there’s a clear trend: fewer people are participating, even though the number of different threads goes up. I don’t think this is because of any decline in Fedora contributors overall — I think it’s that conversations are happening elsewhere.
All of these points ring true for me. I often find myself just avoiding the devel list entirely.
Big threads are … bad, actually
When we have something to talk about, it tends to explode into a big thread. The thing in January with FESCo’s frame pointers decision (https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org/...) is a good example of things going badly.
Most of the conversation was under the subject “Schedule for Tuesday's FESCo Meeting (2023-01-03)”, because everything started as a reply to that. That’s pretty easy to overlook. It’s possible for replies to change the subject when replying, but that can’t be done retroactively, and then isn’t consistent (and it breaks threading in Gmail, too).
Then, things got rather hostile, making it hard to have a reasonable conversation about the issues (both technical and procedural). And then, things went in circles without adding anything new.
This could have all gone a lot better.
And that’s just one example. Take a look back at any mega-thread, and you’ll find similar — and worse. When things get heated, the only way to intervene is by adding more. There are often long subthreads of two people going back and forth on tangents. Then, other conversation branches duplicate that, or refer across. Classical email tools don’t actually handle this kind of thing very well at all. In my experience, it only really works if you keep up with the conversation in almost real time, which has its own problems even when that’s possible.
We’re scattered in actual practice
Devel list may be the center, but we have _hundreds_ of Fedora mailing lists. A dozen or so are reasonably active (Test, Legal, ARM…) but most are inactive or dead. Some are just meeting reminders over and over — for meetings that aren’t even active anymore. It’s easy to make but hard to _unmake_ a mailing list.
For lists that are active, the split is confusing — when should something be on the packaging list rather than devel? What happens when something is related to both Cloud and Server, or Workstation and KDE? One can post to both lists, but if someone replies and isn’t subscribed to both, the conversation gets split.
With “devel” as the main list, conversations about marketing, design, events, and so on don’t really have a central place. (The Mindshare list never really caught on.) That makes these important activities feel even more disconnected and secondary in status — and they shouldn’t be.
Many groups have actually moved away from lists to using tickets for team conversations — both those non-engineering functions and development. Design Team has a mailing list, but mostly for announcements: the work happens in tickets. Workstation largely uses their Pagure tracker. And CoreOS conversations happen almost entirely in tickets on Github.
Tickets are made for tracking specific, actionable tasks, and that kind of tracking is part of why teams use them over mailing lists — but Fedora teams use them for open development conversations too. I think that’s largely a symptom of mailing lists not being enough for what we need. The trackers have media support, editing for typos or updates, reactions for simple agreement, tagging people, and granular subscriptions. They are effectively “off-label use” mini-forums that teams can quietly move to using without the sort of conversation I expect this message to generate.
This is true for EPEL conversations. They tend to happen in IRC/Matrix and Pagure tickets. They also happen on the epel-devel list, but I feel a sense of dread when someone suggests moving a conversation from one of the other places to the list, for all the reasons Matthew mentioned.
Airplane diagram, survivorship bias
Since I’m posting this to devel list, I do expect a lot of push-back. Maybe I’ll be surprised and more of y’all are already with me on this and just waiting for something to happen. But overall I expect a tough crowd.
You’ve probably heard the story about bullet holes in airplanes returning from missions and the accompanying diagram — find it athere https://en.wikipedia.org/wiki/Survivorship_bias if that’s new to you.
The set of remaining regular participants on this list is naturally biased towards those for whom it is working just fine. But, core Fedora development discussion can’t be limited to that ever-shrinking group. Consider who isn’t here. The problems are real, and the trend isn’t in a good direction.
Devel List is too many different things!
We use it for Change discussion (resulting in those not-actually-so-great big threads).
We use it for introductions and onboarding — we’re usually pretty good at that, actually (but it adds to the overall load of following the list). We’re not very consistent, though.
We discuss packaging: guidelines, help on different topics, coordination on specific work. There’s unclear overlap with the packaging list, though, which is a bit confusing.
We talk about higher-level Fedora OS development topics that don’t fit anywhere else. For example, this on installer environment size: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org/... Sometimes, these are FESCo topics (which FESCo heroically struggles to keep here rather than on Pagure tickets…).
Sometimes, people talk about a particular interest in the hopes of forming a Special Interest Group. Often that does result in a new SIG launching — sometimes with a brand-new mailing list which then ends up getting only a few posts ( and a whole lot of spam years later). But this really suffers from the survivorship problem: many people who might be interested will never notice.
We have release process announcements — mostly to devel-announce, but then occasionally replies and discussion here. Blocker bugs and other test-list and QA topics are cross-posted, as part of the release cycle.
And then there are a lot of robo-messages: reports, reminders, etc. These are really valuable to a few people, but add a lot more to wade through for everyone else trying to keep up.
There are certainly others, as well.
In short… there probably shouldn’t just be one thing. But, the cross-posting problem makes it hard to split up as a mailing list.
Enter Discourse
If you’ve talked with me about anything related to any of this in the past ten years, you probably already know that I like Discourse. It’s good software made by cool people. And, it’s entirely open source, with a SaaS business model but with real, usable releases. (No open core, no “open source theoretically but good luck”.) And, we have it in production in Fedora already, at https://discussion.fedoraproject.org, with categories for announcements, user help, project discussion, and social conversation — as well as special categories for dedicated workflows.
In Project Discussion, each different Fedora team can have its own tag, and you can subscribe to those that you’re interested in. Cross-posting is easy: tag a post with multiple teams.
Topics can be renamed, if needed, or split out into side-conversations. The long tangents from these conversations can actually be interesting on their own without distracting from the main topic. Moderation tools allow us to handle posts that are outside of expected Fedora contributor behavior, with varying levels of action as appropriate.
You can use markdown formatting, including images (with easy addition of alt text for people for whom images are a barrier). You can edit your posts to fix typos or correct mistakes. There are polls and lots of other useful features.
And, you can interact with it all by email. That interface isn’t perfect, but it’s way better than any other forum software I’ve seen. (I’ve written a guide: https://discussion.fedoraproject.org/t/25960) At the same time, your individual email address is hidden, so it won’t be a spam magnet (or a target for off-list harassment, a problem we unfortunately have with devel list).
That said, it is web-first software. (Or mobile, if that’s your thing.) That requires some adjustment, I know. I hope opening up a Fedora Discussion tab – or keeping one open — becomes an easy habit.
I went with Discourse over a mailing list when I posted my EPEL 10 proposal back in November. In addition to the aforementioned drawbacks of email, I was specifically interested in being able to use markdown tables and being able to add info to the original post over time. I have enjoyed interacting with that Discourse thread quite a bit, and have no regrets about focusing the conversation there instead of the mailing list.
https://discussion.fedoraproject.org/t/epel-10-proposal/44304
Not just Fedora
There’s a big trend towards Discourse in open source projects overall. Python and Gnome have both migrated entirely from their mailing lists. Ansible is working on it. Plus, there’s Rust, Kubernetes, Nextcloud, Flathub, Grafana, Home Assistant, KDE, and I’m sure many others.
Concrete proposal
I’m not suggesting we shut down devel list next week. And I think we’ll have some mailing lists for quite a long time. But, I think it’s time to start moving some specific things, with the eventual goal of closing every mailing list we can.
First, I’d like to move the Changes discussion. They will still be posted to devel-announce, but responses directed to Project Discussion in a new #changes tag. Ben tells me that this is a FESCo decision, which seems reasonable.
Second, I think other FESCo-related conversations should move. I hope this will reduce the urge to have back-and-forth exchanges in the tickets. For the Fedora Council, I set up a bot which automatically creates a discussion topic when a ticket is filed, leaving the ticket just for votes and recording of outcome. FESCo could use something similar.
Third, I’ll add a tag for general Fedora OS development conversation. Maybe “#devel”, but if someone has more narrow suggestions, I’ll take them.
Fourth, I’d like to update our documentation, process, and expectations for newcomers — say hello on Discussion (and Fedora Chat, if you like) rather than a mailing list. (I’d like to close the Fedora Join list at this point.)
Fifth, all packaging-related discussions (including the separate packaging mailing list). We already have a #package-maintainers tag with some existing discussion.
Sixth, automated posts, as much as we can. These should go to dedicated Workflow categories, where people who want can watch them but where they won’t overwhelm human interactions. People who want can watch them, and it’s easy to quote-reply into a new linked topic in the Project Discussions category.
And finally… shut down the devel list itself. Perhaps at the end of 2023?
We should also shut down all of the little lists that haven’t had anything but spam in the last two years. We could maybe do that sooner. We should stop creating new lists now — we can create new Discussion tags instead.
I expect the announcement lists to stay for the foreseeable future (although we might feed them from Discourse rather than the other way around). Other lists which are patches, commit messages, and other automations might stick around for a while — but really might be better served by a log aggregation and analysis system or something else.
Other teams who want to keep mailing lists can, but I’d like to move those too, and eventually I think we’ll want to shut them down too — or perhaps convert them to announcement lists.
Next steps
I know this is a big change. I’ve been thinking of writing this message for a long time. I’d really like to convince everyone that it’s the right thing — or at least, an acceptable one.
What about specific decisions related to my proposal? For each:
Because altering the Changes process is a FESCo decision, I’ll file a ticket about that shortly.
FESCo moving their own other conversations is, of course, also a FESCo decision.
Assuming the first moves forward, I will create the general #devel tag (or other name we come up with) when I create the #change-proposal tag.
Moving the packaging list is a Packaging Committee decision.
Automated posts can be moved at any time. I can work with the people who own the generation of those reports to figure out a good answer for each.
The outcome for other team lists is up to each team.
And, I think shutting down devel overall is ultimately a Fedora Council decision.
For right now, though: let’s discuss — on the list!
-- Matthew Miller mattdm@fedoraproject.org Fedora Project Leader _______________________________________________ devel mailing list -- devel@lists.fedoraproject.org To unsubscribe send an email to devel-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
Hi Matthew, you say: "We're missing people", and I think, "who?". And who are you going to miss if you move to discourse?
I will be candid, I tried to use forums since the old phpBB times, it never works for me. I have no time to go roaming over forums except if a search engine brings me there.
The mailing list make messages land in my client, on which I am very efficient, therefore I can check all messages once a day, and respond if I find a worthy topic.
Unless this discourse has some great mail bridge (it doesn't) or maybe an rss feed (I do not use those at work, but I guess I could ?) So that I can skim messages on my terms, I think I (and those like me) will be the next "missing people".
Btw I could make exactly the same quote about any forum that Major made for Mailing lists, messy discussions are messy and a forum does not make them easier to follow by any means (perhaps except for those that chose inferior email readers).
All that said, why waste time with this discussion?
Your own post communicates to me (whether you intended it or not) that in the end the thread that will be generated by this post won't matter, because this is just a courtesy post and you already think that the opinion of the "minority of self selected mailing list lovers and dinosaurs" does not matter much.
On Thu, 2023-04-20 at 17:20 -0400, Matthew Miller wrote:
It’s time to transform the Fedora devel list into something new
On Thu, Apr 20, 2023 at 07:21:54PM -0400, Simo Sorce wrote:
Hi Matthew, you say: "We're missing people", and I think, "who?". And who are you going to miss if you move to discourse?
Again and again I have seen this "we're missing people" sentiment be used to justify scrapping "old" workflows, and *not once* has it ever resulted in "more people" coming out of the woodwork that would have happily contributed in the past, but were turned off/away by the need to use archaic email.
(FFS, If we're going to follow this to its logical conclusion, we should just scrap all of this email/discourse/whatever and just move everything to github, or even facebook, as that's clearly where the most numbers of people are. "but no, our custom tooling makes things better for us" is the inevitable pushback, which arguably applies just as much to email-based flows!)
The mailing list make messages land in my client, on which I am very efficient, therefore I can check all messages once a day, and respond if I find a worthy topic.
...and the very nature of Discourse or various other Forums pretty much make this sort of workflow impossible; that is to say you're all but forced to manually poll every site you care about in a way that all but makes automation impossible.
In other words, it's locally optimal for any given site, but is utterly incapable of scaling if you care about more than a small handful of sites.
Calling myself semi-active here would be quite generous, but I can uneqvocibly state that if I have to manually poll a discourse site or whatever, that will be the end of my participating in anything Fedora, except to report bugs via abrt (assuming I don't have to keep logging in for new API keys) I suspect I'm far from the only one in that respect.
Unless this discourse has some great mail bridge (it doesn't) or maybe an rss feed (I do not use those at work, but I guess I could ?) So that I can skim messages on my terms, I think I (and those like me) will be the next "missing people".
RSS doesn't scale for higher volumes unless you're literally polling every few minutes or the feed includes a large number of entries.
Btw I could make exactly the same quote about any forum that Major made for Mailing lists, messy discussions are messy and a forum does not make them easier to follow by any means (perhaps except for those that chose inferior email readers).
Yeah.
Your own post communicates to me (whether you intended it or not) that in the end the thread that will be generated by this post won't matter, because this is just a courtesy post and you already think that the opinion of the "minority of self selected mailing list lovers and dinosaurs" does not matter much.
Unfortunately, I have to agree with this perception. Perhaps it's because I've seen so many other formerly e-mail based communities bifrucate [1] chasing after "engagagement" that never occurs, or even RH/Fedora's near-perfect abysmal record of framing core infrastructure changes like this as a "discussion" when the decision has already been made and will happen no matter what the masses have to say about it.
At the end of the day, distro development, like most other infrastructure, isn't sexy or glamorous, and the humongous effort that goes into it is rarely rewarded with anything other than abuse. "Who cares about distros? I just use Docker containers!"
[1] Splitting into the "core" developers (ie those paid/compensated for participating) and an endless summer of newbs seeking help/support; the middle gets completely hollowed out.
- Solomon
On Thu, Apr 20, 2023 at 8:58 PM Solomon Peachy via devel < devel@lists.fedoraproject.org> wrote:
On Thu, Apr 20, 2023 at 07:21:54PM -0400, Simo Sorce wrote:
Hi Matthew, you say: "We're missing people", and I think, "who?". And who are you going to miss if you move to discourse?
Again and again I have seen this "we're missing people" sentiment be used to justify scrapping "old" workflows, and *not once* has it ever resulted in "more people" coming out of the woodwork that would have happily contributed in the past, but were turned off/away by the need to use archaic email.
(FFS, If we're going to follow this to its logical conclusion, we should just scrap all of this email/discourse/whatever and just move everything to github, or even facebook, as that's clearly where the most numbers of people are. "but no, our custom tooling makes things better for us" is the inevitable pushback, which arguably applies just as much to email-based flows!)
The mailing list make messages land in my client, on which I am very efficient, therefore I can check all messages once a day, and respond if I find a worthy topic.
...and the very nature of Discourse or various other Forums pretty much make this sort of workflow impossible; that is to say you're all but forced to manually poll every site you care about in a way that all but makes automation impossible.
... snip ...
Although I rarely participate, I do read just about everything. And I have to completely agree with Solomon's response.
Moving to a forum based tool will a) have the opposite effect expected, and b) make it harder to follow because of the effort required to be notified, scan and read messages.
- email is push, not poll; so it consumes much less time and bandwidth - email allows 'notification-scan-respond/delete' in basically one (or two) clicks in ~3 seconds of elapsed time (for 'n' messages/topics)
Can you say the same for Discourse?
Am 21.04.2023 um 15:37 schrieb Emmanuel Seyman emmanuel@seyman.fr:
- Fulko Hew [20/04/2023 21:18] :
Can you say the same for Discourse?
... leave a lot to be desired. So much so that
You put it very nicely. I have been desperately trying to follow new posts under the tag #server via email notification. Total failure. I missed everything.
A nice collection of "bells and whistles" just doesn't cut it yet. Currently, it is far too unreliable for systematic and long-term professional activities.
And it is structured in an undercomplex way. Try to find something further back in time. You must have a lot of boredom. A simple grouping by year/month like the mailing lists is very comfortable in comparison.
And I think that's not what discourse was originally designed for. We are now trying to squeeze in something that works "somehow". An 'egg-laying, milk-gearing woolly pig' like with the infamous SAP software in some companies. And the problems that Matthew accurately describes (lengthy discussion, abusive discussion style, etc) cannot be solved by this, in my opinion.
I am very much in favor of modernizing our communication tools. But I don't think this is the way to go (but I don't know anything better, except maybe a more "traditional", more structured forum).
I am very much in favor of modernizing our communication tools. But I
don't think this is the way to go (but I don't know anything better, except maybe a more "traditional", more structured forum).
I'm up for modernizing things and making them more efficient as well, though as Matthew replied to me earlier, that's not his reason for suggesting it.
Having taken some more time to consider this and thinking about the past, I've actually sort of already experienced this when I was a PuppyLinux dev. At the time, we used a forum as their primary means of dev communication. (I think they still do, but I'm not sure) Puppy is a different dev community with many working on their own derivatives of puppy but happily collaborating together with other devs doing their own side thing. While there definitely was an increase in interaction with people, that also came at a cost. Two or three devs having a conversation in a thread would also have to deal with a dozen or so no- devs chiming in on the situation and putting in their two cents. Time and again this would end up being nothing more than a distraction, an argument, or causing a developer to end up going to PMs to discuss the bug they were trying to resolve instead of doing it openly. The users who were jumping in on the thread weren't doing so maliciously, they were actually trying to help, but due to not understanding the issue it ended up doing the opposite.
The problem with mixing general community discussion and development discussion... is you get exactly that as the outcome.
While this did sometimes result in a community member getting more involved in development, I am/was an example of that, it also had the result of causing devs to end up relying on direct emails or PMs to other devs to cut through the chatter of most of the thread being non relevant or wasting time to read through replies that are off topic or unhelpful. And once that communication went private, it was no longer accessible to any other devs to read. This resulted a few times in bugs being fixed in some custom releases but not others, and devs having to re-engineer the same fix because the solution that was found wasn't open. That's less of a problem in Fedora's case because it's not disjointed like Puppy, but it could still happen if there was an issue a dev was dealing with patching source due to some GCC update.
The fact is with a mailing list, someone has to intentionally seek it out and sign up. You dont run into the people casually ending up in developer discussions around esoteric but important engineering discussions.
In the real world I'd relate it to this... A small housing development somewhere, that occasionally has visitors who just drive through to check the place out and look at the neighborhood. Versus When there's a detour on the highway and the entire interstate's traffic ends up getting routed through that same small development to get back onto the interstate later. The first case is one that no one really gets upset at, the latter is one that almost no one is happy with.
I feel like there is a delicate balance between not making communication hard or putting up roadblocks... and making it so easy that the volume becomes a distraction for those that are trying to get code work done. We know what we have now mostly works even though it's not very elegant and has its own issues. Updating tools, encouraging new people to get involved in development are great things, but we dont want to take one step forward which results in two steps back for development overall. Maybe this can be addressed somehow with discourse, but I am clueless as to how. Hopefully those smarter than I can find the right balance for what we should use.
On Fri, Apr 21, 2023 at 10:15 AM Peter Boy pboy@uni-bremen.de wrote:
Am 21.04.2023 um 15:37 schrieb Emmanuel Seyman emmanuel@seyman.fr:
- Fulko Hew [20/04/2023 21:18] :
Can you say the same for Discourse?
... leave a lot to be desired. So much so that
You put it very nicely. I have been desperately trying to follow new posts under the tag #server via email notification. Total failure. I missed everything.
A nice collection of "bells and whistles" just doesn't cut it yet. Currently, it is far too unreliable for systematic and long-term professional activities.
And it is structured in an undercomplex way. Try to find something further back in time. You must have a lot of boredom. A simple grouping by year/month like the mailing lists is very comfortable in comparison.
And I think that's not what discourse was originally designed for. We are now trying to squeeze in something that works "somehow". An 'egg-laying, milk-gearing woolly pig' like with the infamous SAP software in some companies. And the problems that Matthew accurately describes (lengthy discussion, abusive discussion style, etc) cannot be solved by this, in my opinion.
I am very much in favor of modernizing our communication tools. But I don't think this is the way to go (but I don't know anything better, except maybe a more "traditional", more structured forum).
devel mailing list -- devel@lists.fedoraproject.org To unsubscribe send an email to devel-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
On Fri, Apr 21, 2023 at 10:47:24AM -0400, JT wrote:
While there definitely was an increase in interaction with people, that also came at a cost. Two or three devs having a conversation in a thread would also have to deal with a dozen or so no- devs chiming in on the situation and putting in their two cents. Time and again this would end up being nothing more than a distraction, an argument, or causing a developer to end up going to PMs to discuss the bug they were trying to resolve instead of doing it openly.
I don't want to do this prematurely, but we do have some options for handling this. We are (very soon now) going to have sync of FAS groups to Discourse. If we want or need to, we could limit posting in Project Discussion to people who are members of one or more FAS groups.
On Fri, Apr 21, 2023 at 1:15 PM Matthew Miller mattdm@fedoraproject.org wrote:
On Fri, Apr 21, 2023 at 10:47:24AM -0400, JT wrote:
While there definitely was an increase in interaction with people, that also came at a cost. Two or three devs having a conversation in a thread would also have to deal with a dozen or so no- devs chiming in on the situation and putting in their two cents. Time and again this would end
up
being nothing more than a distraction, an argument, or causing a
developer
to end up going to PMs to discuss the bug they were trying to resolve instead of doing it openly.
I don't want to do this prematurely, but we do have some options for handling this. We are (very soon now) going to have sync of FAS groups to Discourse. If we want or need to, we could limit posting in Project Discussion to people who are members of one or more FAS groups.
Oh nice, that'd definitely help prevent issues I dealt with before.
On Fri, Apr 21, 2023 at 04:14:33PM +0200, Peter Boy wrote:
You put it very nicely. I have been desperately trying to follow new posts under the tag #server via email notification. Total failure. I missed everything.
What could have made that go better?
A nice collection of "bells and whistles" just doesn't cut it yet. Currently, it is far too unreliable for systematic and long-term professional activities.
And it is structured in an undercomplex way. Try to find something further back in time. You must have a lot of boredom. A simple grouping by year/month like the mailing lists is very comfortable in comparison.
There isn't (as far as I know) a browsable year/month view, but you can limit searches with before and after: https://discussion.fedoraproject.org/search?q=%23server-wg%20after%3A2021-01...
On 4/21/23 02:57, Solomon Peachy via devel wrote:
On Thu, Apr 20, 2023 at 07:21:54PM -0400, Simo Sorce wrote:
Hi Matthew, you say: "We're missing people", and I think, "who?". And who are you going to miss if you move to discourse?
Again and again I have seen this "we're missing people" sentiment be used to justify scrapping "old" workflows, and *not once* has it ever resulted in "more people" coming out of the woodwork that would have happily contributed in the past, but were turned off/away by the need to use archaic email.
Funny, how from where I am sitting I can not really remember any time we managed to scrape the old workflow at least once. So I wouldn't be able to measure the effectiveness of such an imaginary thing. I can only see how we are unable to do changes and therefore we are always adding things on top of old ones, which of course doesn't make anything easier neither for those who want to change, nor for those who don't.
That's a slight exaggeration of course, but so is your statement. People come to Fedora via many ways. But I doubt any of it starts with e-mail nowadays. And the fact that you don't see newcomers _here_ actually proves the point, isn't it?
(FFS, If we're going to follow this to its logical conclusion, we should just scrap all of this email/discourse/whatever and just move everything to github, or even facebook, as that's clearly where the most numbers of people are. "but no, our custom tooling makes things better for us" is the inevitable pushback, which arguably applies just as much to email-based flows!)
The mailing list make messages land in my client, on which I am very efficient, therefore I can check all messages once a day, and respond if I find a worthy topic.
...and the very nature of Discourse or various other Forums pretty much make this sort of workflow impossible; that is to say you're all but forced to manually poll every site you care about in a way that all but makes automation impossible.
Discourse
1) sends email in a multipart format which has html and plain text(kind of markdown); 2) adds headers which allow you to filter the messages by topic or category on the client-side; 2) allows you to configure notifications for mentions per category or per tag; 3) allows to configure custom searches on the server side, and will notifies you for it.
Here is the example of the mail headers: ----------------------------------------
Message-ID: discourse/post/215862@discussion.fedoraproject.org In-Reply-To: discourse/post/215857@discussion.fedoraproject.org References: discourse/post/215855@discussion.fedoraproject.org discourse/post/215856@discussion.fedoraproject.org discourse/post/215857@discussion.fedoraproject.org List-Unsubscribe: <...> X-Discourse-Post-Id: 215862 X-Discourse-Topic-Id: 81258 X-Discourse-Category: Team Workflows/Fedora Magazine X-Auto-Response-Suppress: All Auto-Submitted: auto-generated Precedence: list List-ID: Fedora Discussion | Team Workflows Fedora Magazine <fedora-magazine.team-workflows.discussion.fedoraproject.org> List-Archive: https://discussion.fedoraproject.org/t/article-proposal-using-btrfs-to-upgra... Feedback-ID: fedoraproject:user_posted:discoursemail
----------------------------------------
It is not perfect, and it requires some effort. But I don't see why it is impossible.
In other words, it's locally optimal for any given site, but is utterly incapable of scaling if you care about more than a small handful of sites.
Calling myself semi-active here would be quite generous, but I can uneqvocibly state that if I have to manually poll a discourse site or whatever, that will be the end of my participating in anything Fedora, except to report bugs via abrt (assuming I don't have to keep logging in for new API keys) I suspect I'm far from the only one in that respect.
Let's not get into a "who would you miss more" competition and work on a solution which actually helps us to bridge the gap and allows us to compromise between different use cases.
Unless this discourse has some great mail bridge (it doesn't) or maybe an rss feed (I do not use those at work, but I guess I could ?) So that I can skim messages on my terms, I think I (and those like me) will be the next "missing people".
RSS doesn't scale for higher volumes unless you're literally polling every few minutes or the feed includes a large number of entries.
Btw I could make exactly the same quote about any forum that Major made for Mailing lists, messy discussions are messy and a forum does not make them easier to follow by any means (perhaps except for those that chose inferior email readers).
Yeah.
Your own post communicates to me (whether you intended it or not) that in the end the thread that will be generated by this post won't matter, because this is just a courtesy post and you already think that the opinion of the "minority of self selected mailing list lovers and dinosaurs" does not matter much.
Unfortunately, I have to agree with this perception. Perhaps it's because I've seen so many other formerly e-mail based communities bifrucate [1] chasing after "engagagement" that never occurs, or even RH/Fedora's near-perfect abysmal record of framing core infrastructure changes like this as a "discussion" when the decision has already been made and will happen no matter what the masses have to say about it.
At the end of the day, distro development, like most other infrastructure, isn't sexy or glamorous, and the humongous effort that goes into it is rarely rewarded with anything other than abuse. "Who cares about distros? I just use Docker containers!"
In all seriousness, I would advise you to hang out at the current discussion.fedoraproject.org and feel the vibe a bit.
Distributions _are_ cool and sexy. And people have ideas and interest in them. Some of them are totally wrong and misplaced, some may be very old, and some are better. But that's how it should be.
It seems you feel like you are cornered, but it is you who put yourself in the corner by ignoring the part of the community, which actually can and wants to support you.
[1] Splitting into the "core" developers (ie those paid/compensated for participating) and an endless summer of newbs seeking help/support; the middle gets completely hollowed out.
- Solomon
Just a slight addition about "archaic email" and related comments:
Email and its capability for being used in conjunction with OpenPGP ensures two major institutions in kernel development and elsewhere: "Trusting the developers, not infrastructure" [1], and, assume "any part of the infrastructure can be compromised at any time" [1]. This avoids single points of failure, and complements the chain of trust.
I am not sure if Discourse is capable to be used in conjunction with OpenPGP if it reformats contents or if it removes attachments (maybe someone knows?). This leads to the possibility that discourse introduces a single point of failure (or, single point of vulnerability), which breaks the above institutions.
Having said that, as far as I follow our devel mailing list, I think the argument above is of minor relevance, because I think this mailing list is not used to forward code or to do reviews. Signatures seem to be not omnipresent at the moment anyway.
However, I just wanted to remind that the issue is a little more complex than just assuming "email is old and has to be replaced by modern": there is another consideration, too. And we have to be aware that if discourse does not support OpenPGP signatures practically, we loose the possibility to ensure "security of integrity" in the mailing list in cases WHEN it is necessary - IF there are such cases (which I cannot determine).
Just some thoughts :)
[1] https://www.kernel.org/doc/html/latest/process/maintainer-pgp-guide.html
Chris
On 4/21/23 11:42, Aleksandra Fedorova wrote:
On 4/21/23 02:57, Solomon Peachy via devel wrote:
On Thu, Apr 20, 2023 at 07:21:54PM -0400, Simo Sorce wrote:
Hi Matthew, you say: "We're missing people", and I think, "who?". And who are you going to miss if you move to discourse?
Again and again I have seen this "we're missing people" sentiment be used to justify scrapping "old" workflows, and *not once* has it ever resulted in "more people" coming out of the woodwork that would have happily contributed in the past, but were turned off/away by the need to use archaic email.
Funny, how from where I am sitting I can not really remember any time we managed to scrape the old workflow at least once. So I wouldn't be able to measure the effectiveness of such an imaginary thing. I can only see how we are unable to do changes and therefore we are always adding things on top of old ones, which of course doesn't make anything easier neither for those who want to change, nor for those who don't.
That's a slight exaggeration of course, but so is your statement. People come to Fedora via many ways. But I doubt any of it starts with e-mail nowadays. And the fact that you don't see newcomers _here_ actually proves the point, isn't it?
(FFS, If we're going to follow this to its logical conclusion, we should just scrap all of this email/discourse/whatever and just move everything to github, or even facebook, as that's clearly where the most numbers of people are. "but no, our custom tooling makes things better for us" is the inevitable pushback, which arguably applies just as much to email-based flows!)
The mailing list make messages land in my client, on which I am very efficient, therefore I can check all messages once a day, and respond if I find a worthy topic.
...and the very nature of Discourse or various other Forums pretty much make this sort of workflow impossible; that is to say you're all but forced to manually poll every site you care about in a way that all but makes automation impossible.
Discourse
- sends email in a multipart format which has html and plain
text(kind of markdown); 2) adds headers which allow you to filter the messages by topic or category on the client-side; 2) allows you to configure notifications for mentions per category or per tag; 3) allows to configure custom searches on the server side, and will notifies you for it.
Here is the example of the mail headers:
Message-ID: discourse/post/215862@discussion.fedoraproject.org In-Reply-To: discourse/post/215857@discussion.fedoraproject.org References: discourse/post/215855@discussion.fedoraproject.org discourse/post/215856@discussion.fedoraproject.org discourse/post/215857@discussion.fedoraproject.org List-Unsubscribe: <...> X-Discourse-Post-Id: 215862 X-Discourse-Topic-Id: 81258 X-Discourse-Category: Team Workflows/Fedora Magazine X-Auto-Response-Suppress: All Auto-Submitted: auto-generated Precedence: list List-ID: Fedora Discussion | Team Workflows Fedora Magazine <fedora-magazine.team-workflows.discussion.fedoraproject.org> List-Archive: https://discussion.fedoraproject.org/t/article-proposal-using-btrfs-to-upgra... Feedback-ID: fedoraproject:user_posted:discoursemail
It is not perfect, and it requires some effort. But I don't see why it is impossible.
In other words, it's locally optimal for any given site, but is utterly incapable of scaling if you care about more than a small handful of sites.
Calling myself semi-active here would be quite generous, but I can uneqvocibly state that if I have to manually poll a discourse site or whatever, that will be the end of my participating in anything Fedora, except to report bugs via abrt (assuming I don't have to keep logging in for new API keys) I suspect I'm far from the only one in that respect.
Let's not get into a "who would you miss more" competition and work on a solution which actually helps us to bridge the gap and allows us to compromise between different use cases.
Unless this discourse has some great mail bridge (it doesn't) or maybe an rss feed (I do not use those at work, but I guess I could ?) So that I can skim messages on my terms, I think I (and those like me) will be the next "missing people".
RSS doesn't scale for higher volumes unless you're literally polling every few minutes or the feed includes a large number of entries.
Btw I could make exactly the same quote about any forum that Major made for Mailing lists, messy discussions are messy and a forum does not make them easier to follow by any means (perhaps except for those that chose inferior email readers).
Yeah.
Your own post communicates to me (whether you intended it or not) that in the end the thread that will be generated by this post won't matter, because this is just a courtesy post and you already think that the opinion of the "minority of self selected mailing list lovers and dinosaurs" does not matter much.
Unfortunately, I have to agree with this perception. Perhaps it's because I've seen so many other formerly e-mail based communities bifrucate [1] chasing after "engagagement" that never occurs, or even RH/Fedora's near-perfect abysmal record of framing core infrastructure changes like this as a "discussion" when the decision has already been made and will happen no matter what the masses have to say about it.
At the end of the day, distro development, like most other infrastructure, isn't sexy or glamorous, and the humongous effort that goes into it is rarely rewarded with anything other than abuse. "Who cares about distros? I just use Docker containers!"
In all seriousness, I would advise you to hang out at the current discussion.fedoraproject.org and feel the vibe a bit.
Distributions _are_ cool and sexy. And people have ideas and interest in them. Some of them are totally wrong and misplaced, some may be very old, and some are better. But that's how it should be.
It seems you feel like you are cornered, but it is you who put yourself in the corner by ignoring the part of the community, which actually can and wants to support you.
[1] Splitting into the "core" developers (ie those paid/compensated for participating) and an endless summer of newbs seeking help/support; the middle gets completely hollowed out.
- Solomon
On 4/21/23 15:25, Christopher Klooz wrote:
Just a slight addition about "archaic email" and related comments:
Email and its capability for being used in conjunction with OpenPGP ensures two major institutions in kernel development and elsewhere: "Trusting the developers, not infrastructure" [1], and, assume "any part of the infrastructure can be compromised at any time" [1]. This avoids single points of failure, and complements the chain of trust.
I am not sure if Discourse is capable to be used in conjunction with OpenPGP if it reformats contents or if it removes attachments (maybe someone knows?). This leads to the possibility that discourse introduces a single point of failure (or, single point of vulnerability), which breaks the above institutions.
Having said that, as far as I follow our devel mailing list, I think the argument above is of minor relevance, because I think this mailing list is not used to forward code or to do reviews. Signatures seem to be not omnipresent at the moment anyway.
From security or impersonation point of view our current mailing list is actually the worst. Both Matrix and Discourse are at least tied to FAS account. And while it can be considered a single point of failure, it is at least the one which exists and is properly maintained by the project.
We had the issue with impersonation over e-mail before, and that was not nice.
However, I just wanted to remind that the issue is a little more complex than just assuming "email is old and has to be replaced by modern": there is another consideration, too. And we have to be aware that if discourse does not support OpenPGP signatures practically, we loose the possibility to ensure "security of integrity" in the mailing list in cases WHEN it is necessary - IF there are such cases (which I cannot determine).
I think we really try hard to not oversimplify the conversation to the point of "old" vs "new", or "us" vs "them" approach, though many of the replies in this thread are pulling us into that direction.
Matthew's mail in my opinion does a good job to highlight that there is no single "we want a new shiny thing for newbies" driver behind the switch. There are multiple reasons for it. And making discussions more secure and better maintained is on that list too.
And like, hey, e-mail is a still a thing. Use it where you need it, and where it fits. There is no fight against the technology.
But for this particular purpose within this particular environment the mailing list just doesn't work(*), and we see it.
(*) Works = provides shared space where old and new Fedora contributors can discuss changes and other project-related topics in a collaborative way to advance the project.
This is the problem which we must solve. And it won't go away on its own if just wait for it.
Again, the goal is not to fight against Fedora contributors using the e-mail technology. The goal is to find a solution.
And if the requirement for that solution is to improve the Discourse mail interface, can we at least try to look into it with open mind and actually list what needs to be done to make it work.
We are a group of FOSS developers using FOSS tools, and we have a year long plan to make the tool working for us and everyone else.
Let's maybe work on it?
Just some thoughts :)
[1] https://www.kernel.org/doc/html/latest/process/maintainer-pgp-guide.html
Chris
On 4/21/23 16:30, Aleksandra Fedorova wrote:
On 4/21/23 15:25, Christopher Klooz wrote:
Just a slight addition about "archaic email" and related comments:
Email and its capability for being used in conjunction with OpenPGP ensures two major institutions in kernel development and elsewhere: "Trusting the developers, not infrastructure" [1], and, assume "any part of the infrastructure can be compromised at any time" [1]. This avoids single points of failure, and complements the chain of trust.
I am not sure if Discourse is capable to be used in conjunction with OpenPGP if it reformats contents or if it removes attachments (maybe someone knows?). This leads to the possibility that discourse introduces a single point of failure (or, single point of vulnerability), which breaks the above institutions.
Having said that, as far as I follow our devel mailing list, I think the argument above is of minor relevance, because I think this mailing list is not used to forward code or to do reviews. Signatures seem to be not omnipresent at the moment anyway.
From security or impersonation point of view our current mailing list is actually the worst. Both Matrix and Discourse are at least tied to FAS account. And while it can be considered a single point of failure, it is at least the one which exists and is properly maintained by the project.
The FAS account is useless if one has access to the infra, or if the latter has vulnerabilities (which can be social and technical). Misconfigurations also occur in complex infra. That's the point of avoiding single points of failure. If one uses OpenPGP and if people verify it, it is not relevant if the infra itself is the "worst" or not, because no one needs to trust it anyway (that's the point in the kernel mailing lists). By default, without ensuring integrity, every email-based mailing list that is used in Linux realms (and at all) falls in the "worst" category because of the concept/architecture of email.
Again, this does not mean that discourse is not suitable for us. Given what I see on the mailing lists, our mailing list contents seem to be not relevant for integrity, and mostly not signed at all.
I just read some comments where I had the perception that they are partly assuming things to be simpler than they are. There are reasons for traditional email mailing lists in some circumstances, they are not "generally obsolete", but this does not mean that this applies to our mailing lists.
Given what I see and where I am present in the mailing lists, I would be +1 for discourse. But we still have to consider and put forward all points.
So I think we are on the same page, I just added a point that has to be considered in advance: do we have >=1 mailing lists that have a need for independent "security of integrity"? I guess the answer is no, we do not have >=1. But I do not know all of our mailing lists and for what they are used.
We had the issue with impersonation over e-mail before, and that was not nice.
However, I just wanted to remind that the issue is a little more complex than just assuming "email is old and has to be replaced by modern": there is another consideration, too. And we have to be aware that if discourse does not support OpenPGP signatures practically, we loose the possibility to ensure "security of integrity" in the mailing list in cases WHEN it is necessary - IF there are such cases (which I cannot determine).
I think we really try hard to not oversimplify the conversation to the point of "old" vs "new", or "us" vs "them" approach, though many of the replies in this thread are pulling us into that direction.
Matthew's mail in my opinion does a good job to highlight that there is no single "we want a new shiny thing for newbies" driver behind the switch. There are multiple reasons for it. And making discussions more secure and better maintained is on that list too.
And like, hey, e-mail is a still a thing. Use it where you need it, and where it fits. There is no fight against the technology.
But for this particular purpose within this particular environment the mailing list just doesn't work(*), and we see it.
(*) Works = provides shared space where old and new Fedora contributors can discuss changes and other project-related topics in a collaborative way to advance the project.
This is the problem which we must solve. And it won't go away on its own if just wait for it.
Again, the goal is not to fight against Fedora contributors using the e-mail technology. The goal is to find a solution.
And if the requirement for that solution is to improve the Discourse mail interface, can we at least try to look into it with open mind and actually list what needs to be done to make it work.
We are a group of FOSS developers using FOSS tools, and we have a year long plan to make the tool working for us and everyone else.
Let's maybe work on it?
Just some thoughts :)
[1] https://www.kernel.org/doc/html/latest/process/maintainer-pgp-guide.html
Chris
On 4/21/23 17:27, Daniel Alley wrote:
If one uses OpenPGP and if people verify it
As you mention, that's a big "if"
Absolutely, and if the majority does not verify in the devel mailing list, it is clearly an indicator that this type of security is not relevant here ;) But finally, I am not sufficiently involved in all of our mailing lists to know.
devel mailing list -- devel@lists.fedoraproject.org To unsubscribe send an email to devel-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
On Fri, Apr 21, 2023 at 11:42:20AM +0200, Aleksandra Fedorova wrote:
That's a slight exaggeration of course, but so is your statement. People come to Fedora via many ways. But I doubt any of it starts with e-mail nowadays. And the fact that you don't see newcomers _here_ actually proves the point, isn't it?
Correlation is not causation.
Distro building isn't "fun", or sexy. There are much more immediately (and fiscally) rewarding things for "newcomers" to mess with.
Let's not get into a "who would you miss more" competition and work on a solution which actually helps us to bridge the gap and allows us to compromise between different use cases.
Oh, I know my contributions here are miniscule, but my point is that we'd lose a ton of voices that collectively represent a ton of experience and use cases.
In all seriousness, I would advise you to hang out at the current discussion.fedoraproject.org and feel the vibe a bit.
You kinda just demonstrated my point -- "hanging out at <new web site> and feel the vibe" is going to take time, attention, and distruption. My total available time/attention is fixed, which means this "hanging out" will have to come at the cost of something else that frankly matters _more_. And I won't personally gain anything from the effort -- at _best_ it will break even vs what I have now.
Distributions _are_ cool and sexy. And people have ideas and interest in them. Some of them are totally wrong and misplaced, some may be very old, and some are better. But that's how it should be.
I'm sorry, the numbers are _not_ on your side here, not just with Fedora itself but the bigger picture of distributions in general. It's not "email" that is keeping folks away, it's the nature of the work. And _work_ it absolutely is.
(And I say this as someone who has spent most of the past couple of decades working on low-level infrastructure-type stuff. Sure, I find it fun/enjoyable but I freely acknowledge I am several standard deviations from the mean)
It seems you feel like you are cornered, but it is you who put yourself in the corner by ignoring the part of the community, which actually can and wants to support you.
By that same token bananas could make themselves more appealing to people that like oranges if they'd only be more orange-like.
This isn't me "putting myself in a corner", it's Fedora moving the tent that I was underneath and expecting me to move with it. (Because they either don't understand _why_ anyone wouldn't want to move, or understand, and do it anyway. I can actually respect the latter position, even if I think it's not going to yield the expected results. But I think this is yet another case of the former)
But whatever, I won't lose any sleep over this. As I already mentioned, I have plenty of other things to do, both in the F/OSS world and in (gasp) meatspace where I won't have to look at yet another screen.
[1] Splitting into the "core" developers (ie those paid/compensated for participating) and an endless summer of newbs seeking help/support; the middle gets completely hollowed out.
FWIW, I stand by this analogy.
- Solomon
Am 21.04.2023 um 16:15 schrieb Solomon Peachy via devel devel@lists.fedoraproject.org:
You kinda just demonstrated my point -- "hanging out at <new web site> and feel the vibe" is going to take time, attention, and distruption. My total available time/attention is fixed, which means this "hanging out" will have to come at the cost of something ...
+++…+++1 :-)
On 4/21/23 16:15, Solomon Peachy via devel wrote:
On Fri, Apr 21, 2023 at 11:42:20AM +0200, Aleksandra Fedorova wrote:
That's a slight exaggeration of course, but so is your statement. People come to Fedora via many ways. But I doubt any of it starts with e-mail nowadays. And the fact that you don't see newcomers _here_ actually proves the point, isn't it?
Correlation is not causation.
Distro building isn't "fun", or sexy. There are much more immediately (and fiscally) rewarding things for "newcomers" to mess with.
Let's not get into a "who would you miss more" competition and work on a solution which actually helps us to bridge the gap and allows us to compromise between different use cases.
Oh, I know my contributions here are miniscule, but my point is that we'd lose a ton of voices that collectively represent a ton of experience and use cases.
In all seriousness, I would advise you to hang out at the current discussion.fedoraproject.org and feel the vibe a bit.
You kinda just demonstrated my point -- "hanging out at <new web site> and feel the vibe" is going to take time, attention, and distruption. My total available time/attention is fixed, which means this "hanging out" will have to come at the cost of something else that frankly matters _more_. And I won't personally gain anything from the effort -- at _best_ it will break even vs what I have now.
No, not really. I advised you to look into it, because you may actually get more from it personally, than you currently expect. I haven't proposed it to become a required Fedora activity, I gave you the unsolicited advice. Which maybe I shouldn't have.
Distributions _are_ cool and sexy. And people have ideas and interest in them. Some of them are totally wrong and misplaced, some may be very old, and some are better. But that's how it should be.
I'm sorry, the numbers are _not_ on your side here, not just with Fedora itself but the bigger picture of distributions in general. It's not "email" that is keeping folks away, it's the nature of the work. And _work_ it absolutely is.
And now you try to tell me that "coolness" is defined by numbers. Sorry, I don't agree. I don't need everyone in the world to work on Linux distribution. I don't want majority of people to work on Linux distribution. I don't even need "more than in Ubuntu" people to work on Fedora Linux distribution. That is not the point at all, and it is not what makes things cool.
But with my Fedora Ambassador hat on I can tell you that the problem we see right now is not that we don't have people coming to Fedora. We have a problem helping people to connect to where the work is happening in a way that they can contribute.
And this includes both mentoring them to be able to contribute, but also accepting the fact that new people can bring new ideas, and we should provide them space to work on them and not just expect them to follow and do what they were told to do.
(And I say this as someone who has spent most of the past couple of decades working on low-level infrastructure-type stuff. Sure, I find it fun/enjoyable but I freely acknowledge I am several standard deviations from the mean)
It seems you feel like you are cornered, but it is you who put yourself in the corner by ignoring the part of the community, which actually can and wants to support you.
By that same token bananas could make themselves more appealing to people that like oranges if they'd only be more orange-like.
This isn't me "putting myself in a corner", it's Fedora moving the tent that I was underneath and expecting me to move with it. (Because they either don't understand _why_ anyone wouldn't want to move, or understand, and do it anyway. I can actually respect the latter position, even if I think it's not going to yield the expected results. But I think this is yet another case of the former)
But whatever, I won't lose any sleep over this. As I already mentioned, I have plenty of other things to do, both in the F/OSS world and in (gasp) meatspace where I won't have to look at yet another screen.
[1] Splitting into the "core" developers (ie those paid/compensated for participating) and an endless summer of newbs seeking help/support; the middle gets completely hollowed out.
FWIW, I stand by this analogy.
The middle doesn't magically appear out of nowhere. It appears when you build paths for newbies to grow into it. And it it sort of responsibility of the current middle to grow the next one.
We think the communication channels change is one of the initiatives which helps with that. Mentorship is the other.
There are currently three initiatives under Sustainable Community objective and everyone's feedback is welcome:
https://discussion.fedoraproject.org/t/fedora-strategy-2028-focus-area-revie...
- Solomon
devel mailing list -- devel@lists.fedoraproject.org To unsubscribe send an email to devel-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
But with my Fedora Ambassador hat on I can tell you that the problem
we see right now is not that we don't have people coming to Fedora. We have a problem helping people to connect to where the work is happening in a way that they can contribute.
And this includes both mentoring them to be able to contribute, but also
accepting the fact that new people can bring new ideas, and we should provide them space to work on them and not just expect them to follow and do what they were told to do.
So I'm interested by what you bring up here. Have you run into situations where someone wanted to contribute to development but was unwilling to use a mailing list? With a community as big as Fedora and with a multitude of ways that people can contribute, I'm curious what the roadblocks you are seeing for people wanting to get into development. I can completely understand if someone wants to join mindshare, D&I, outreachy, or docs, etc... that they might find a mailinglist to cumbersome to work with. Have you run into sitautions where people wanted to get involved in development but were having issues with a mailing list?
I'm involved in many things FOSS related, from development, to outreach, to media production, etc. I've found that in those different 'ecosystems' for lack of a better word, there are some tools that are more effective than others. I think that there are some things that happen on the mailing list that might certainly benefit from moving to discourse, but I also feel that there are some things that benefit from the current system until something new is designed.
On Fri, Apr 21, 2023 at 11:07 AM Aleksandra Fedorova alpha@bookwar.info wrote:
On 4/21/23 16:15, Solomon Peachy via devel wrote:
On Fri, Apr 21, 2023 at 11:42:20AM +0200, Aleksandra Fedorova wrote:
That's a slight exaggeration of course, but so is your statement. People come to Fedora via many ways. But I doubt any of it starts with e-mail nowadays. And the fact that you don't see newcomers _here_ actually
proves
the point, isn't it?
Correlation is not causation.
Distro building isn't "fun", or sexy. There are much more immediately (and fiscally) rewarding things for "newcomers" to mess with.
Let's not get into a "who would you miss more" competition and work on a solution which actually helps us to bridge the gap and allows us to compromise between different use cases.
Oh, I know my contributions here are miniscule, but my point is that we'd lose a ton of voices that collectively represent a ton of experience and use cases.
In all seriousness, I would advise you to hang out at the current discussion.fedoraproject.org and feel the vibe a bit.
You kinda just demonstrated my point -- "hanging out at <new web site> and feel the vibe" is going to take time, attention, and distruption. My total available time/attention is fixed, which means this "hanging out" will have to come at the cost of something else that frankly matters _more_. And I won't personally gain anything from the effort -- at _best_ it will break even vs what I have now.
No, not really. I advised you to look into it, because you may actually get more from it personally, than you currently expect. I haven't proposed it to become a required Fedora activity, I gave you the unsolicited advice. Which maybe I shouldn't have.
Distributions _are_ cool and sexy. And people have ideas and interest in them. Some of them are totally wrong and misplaced, some may be very
old,
and some are better. But that's how it should be.
I'm sorry, the numbers are _not_ on your side here, not just with Fedora itself but the bigger picture of distributions in general. It's not "email" that is keeping folks away, it's the nature of the work. And _work_ it absolutely is.
And now you try to tell me that "coolness" is defined by numbers. Sorry, I don't agree. I don't need everyone in the world to work on Linux distribution. I don't want majority of people to work on Linux distribution. I don't even need "more than in Ubuntu" people to work on Fedora Linux distribution. That is not the point at all, and it is not what makes things cool.
But with my Fedora Ambassador hat on I can tell you that the problem we see right now is not that we don't have people coming to Fedora. We have a problem helping people to connect to where the work is happening in a way that they can contribute.
And this includes both mentoring them to be able to contribute, but also accepting the fact that new people can bring new ideas, and we should provide them space to work on them and not just expect them to follow and do what they were told to do.
(And I say this as someone who has spent most of the past couple of decades working on low-level infrastructure-type stuff. Sure, I find it fun/enjoyable but I freely acknowledge I am several standard deviations from the mean)
It seems you feel like you are cornered, but it is you who put yourself
in
the corner by ignoring the part of the community, which actually can and wants to support you.
By that same token bananas could make themselves more appealing to people that like oranges if they'd only be more orange-like.
This isn't me "putting myself in a corner", it's Fedora moving the tent that I was underneath and expecting me to move with it. (Because they either don't understand _why_ anyone wouldn't want to move, or understand, and do it anyway. I can actually respect the latter position, even if I think it's not going to yield the expected results. But I think this is yet another case of the former)
But whatever, I won't lose any sleep over this. As I already mentioned, I have plenty of other things to do, both in the F/OSS world and in (gasp) meatspace where I won't have to look at yet another screen.
[1] Splitting into the "core" developers (ie those paid/compensated for participating) and an endless summer of newbs seeking
help/support;
the middle gets completely hollowed out.FWIW, I stand by this analogy.
The middle doesn't magically appear out of nowhere. It appears when you build paths for newbies to grow into it. And it it sort of responsibility of the current middle to grow the next one.
We think the communication channels change is one of the initiatives which helps with that. Mentorship is the other.
There are currently three initiatives under Sustainable Community objective and everyone's feedback is welcome:
https://discussion.fedoraproject.org/t/fedora-strategy-2028-focus-area-revie...
- Solomon
devel mailing list -- devel@lists.fedoraproject.org To unsubscribe send an email to devel-leave@lists.fedoraproject.org Fedora Code of Conduct:
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives:
https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org
Do not reply to spam, report it:
https://pagure.io/fedora-infrastructure/new_issue _______________________________________________ devel mailing list -- devel@lists.fedoraproject.org To unsubscribe send an email to devel-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
On Fri, Apr 21, 2023 at 11:54:09AM -0400, JT wrote:
So I'm interested by what you bring up here. Have you run into situations where someone wanted to contribute to development but was unwilling to use a mailing list? With a community as big as Fedora and with a multitude of
I talk to a lot of people about getting involved in Fedora. "Sign up for this mailing list and introduce yourself" is a big drop-out point.
On 4/21/23 22:07, Matthew Miller wrote:
On Fri, Apr 21, 2023 at 11:54:09AM -0400, JT wrote:
So I'm interested by what you bring up here. Have you run into situations where someone wanted to contribute to development but was unwilling to use a mailing list? With a community as big as Fedora and with a multitude of
I talk to a lot of people about getting involved in Fedora. "Sign up for this mailing list and introduce yourself" is a big drop-out point.
I would think the latter to be the bigger obstacle of the two.
- Panu -
On Fri, Apr 21, 2023 at 11:54:09AM -0400, JT wrote:
But with my Fedora Ambassador hat on I can tell you that the problem
we see right now is not that we don't have people coming to Fedora. We have a problem helping people to connect to where the work is happening in a way that they can contribute.
And this includes both mentoring them to be able to contribute, but also
accepting the fact that new people can bring new ideas, and we should provide them space to work on them and not just expect them to follow and do what they were told to do.
[^^^^^^ side note about the above ^^^^^^^ Your mail client breaks quoting: it only inserts ">" on the first line, and not the later lines of the quote. This makes your mails harder to read than they should be. The same is true in your other mails, it's not a one-off thing.]
Have you run into situations where someone wanted to contribute to development but was unwilling to use a mailing list? With a community as big as Fedora and with a multitude of ways that people can contribute, I'm curious what the roadblocks you are seeing for people wanting to get into development. I can completely understand if someone wants to join mindshare, D&I, outreachy, or docs, etc... that they might find a mailinglist to cumbersome to work with. Have you run into sitautions where people wanted to get involved in development but were having issues with a mailing list?
All those things are *development*. Without infra and docs and people and communication, the remaining development "core" would mean very little. And to make the project effective we absolutely need to cooperate and coordinate across all those teams.
So if we have whole groups of people who find [some communication medium] tedious, it is a very strong argument against that medium. (I'm still reading the thread and haven't made up my mind, but if anything, this part of the discussion is a strong argument *for* discourse, because there are clear problems with the mailing list approach and it's easier for long-time contributors like you and me to adjust than for newcomers.)
Zbyszek
On Sat, Apr 22, 2023 at 04:05:41PM +0000, Zbigniew Jędrzejewski-Szmek wrote:
And this includes both mentoring them to be able to contribute, but also
accepting the fact that new people can bring new ideas, and we should provide them space to work on them and not just expect them to follow and do what they were told to do.
[^^^^^^ side note about the above ^^^^^^^ Your mail client breaks quoting: it only inserts ">" on the first line, and not the later lines of the quote. This makes your mails harder to read than they should be. The same is true in your other mails, it's not a one-off thing.]
This is probably Gmail's web client. It does this, and other odd things that make it very frustrating to use for anything but "top-post, quote-everything".
On Fri, Apr 21, 2023 at 05:07:28PM +0200, Aleksandra Fedorova wrote:
No, not really. I advised you to look into it, because you may actually get more from it personally, than you currently expect. I haven't proposed it to become a required Fedora activity, I gave you the unsolicited advice. Which maybe I shouldn't have.
What kicked off this thread was a proposal to make "checking out" (and then actually utilizing) discussion.f.o a *required* activity to participate in Fedora development.
And now you try to tell me that "coolness" is defined by numbers. Sorry, I don't agree. I don't need everyone in the world to work on Linux distribution. I don't want majority of people to work on Linux distribution. I don't even need "more than in Ubuntu" people to work on Fedora Linux distribution. That is not the point at all, and it is not what makes things cool.
Numbers/engagement/etc was one of the justifications behind this proposal to ditch mailing lists, and my point was that "numbers" was a pretty worthless metric because of the very reasons you just listed.
That said, I don't think anyone can reasonably disagree that the "coolness" factor strongly affects the numbers, because people follow the funding. Distros, like most(all) other infrastructure, were never "cool" in that big-picture context. I think that's a good thing, as it means those participating are more likely to have a genuine passion for it.
But with my Fedora Ambassador hat on I can tell you that the problem we see right now is not that we don't have people coming to Fedora. We have a problem helping people to connect to where the work is happening in a way that they can contribute.
Of course. But this also goes for those that are _already_ contributing.
And this includes both mentoring them to be able to contribute, but also accepting the fact that new people can bring new ideas, and we should provide them space to work on them and not just expect them to follow and do what they were told to do.
This also goes for those that are already contributing. It's not just about the new people.
The middle doesn't magically appear out of nowhere. It appears when you build paths for newbies to grow into it. And it it sort of responsibility of the current middle to grow the next one.
I disagree; it's the _responsibility_ of the *leaders* to grow (and sustain!) the middle. Ultimately this comes down to defining the sort of culture the community is expected to have -- and the tools must be chosen to support/enable that culture.
(It's not the middle's responsibility because they are acting mostly individually and mostly powerless; their main power comes from the ballot box, except in rare situations where someone rises up through sheer volume of meritocrous contributions)
We think the communication channels change is one of the initiatives which helps with that. Mentorship is the other.
Mentorship is necessary (from leaders to the middle, and the middle to the newbs) but doesn't scale well due to the mutual time committment necessary.
- Solomon
Am 21.04.23 um 16:15 schrieb Solomon Peachy via devel:
On Fri, Apr 21, 2023 at 11:42:20AM +0200, Aleksandra Fedorova wrote:
In all seriousness, I would advise you to hang out at the current discussion.fedoraproject.org and feel the vibe a bit.
You kinda just demonstrated my point -- "hanging out at <new web site> and feel the vibe" is going to take time, attention, and distruption.
Absolutely.
The proposal is harmful to Fedora and likely based on the distorted view of somebody lacking of expericence on practical distribution work.
Ralf
On Fri, 21 Apr 2023 at 14:07, Ralf Corsépius rc040203@freenet.de wrote:
Am 21.04.23 um 16:15 schrieb Solomon Peachy via devel:
On Fri, Apr 21, 2023 at 11:42:20AM +0200, Aleksandra Fedorova wrote:
In all seriousness, I would advise you to hang out at the current discussion.fedoraproject.org and feel the vibe a bit.
You kinda just demonstrated my point -- "hanging out at <new web site> and feel the vibe" is going to take time, attention, and distruption.
Absolutely.
The proposal is harmful to Fedora and likely based on the distorted view of somebody lacking of expericence on practical distribution work.
I am going to say that comment is Wrong in multiple ways. Both Matthew and Aleksandra have done deep and practical distribution work. You may not like their proposal, and it may not be how you or I find communication to be useful, BUT making comments about a person's abilities is out of line.
Ralf _______________________________________________ devel mailing list -- devel@lists.fedoraproject.org To unsubscribe send an email to devel-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
On 4/21/23 02:21, Simo Sorce wrote:
Hi Matthew, you say: "We're missing people", and I think, "who?". And who are you going to miss if you move to discourse?
I will be candid, I tried to use forums since the old phpBB times, it never works for me. I have no time to go roaming over forums except if a search engine brings me there.
The mailing list make messages land in my client, on which I am very efficient, therefore I can check all messages once a day, and respond if I find a worthy topic.
Unless this discourse has some great mail bridge (it doesn't) or maybe an rss feed (I do not use those at work, but I guess I could ?) So that I can skim messages on my terms, I think I (and those like me) will be the next "missing people".
Ditto.
I actually quite like Discourse - for a forum software - from experience related to various freetime activities.
However, Discourse replacing mailing lists WILL be the end of habitually skimming through everything that goes on devel (and a whole bunch of other lists) to spot issues that might be of my concern. The result will be in me being considerably less aware of what goes around in Fedora, rpm related or not.
Just FWIW.
- Panu -
On Fri, 21 Apr 2023 10:31:04 +0300 Panu Matilainen pmatilai@redhat.com wrote:
On 4/21/23 02:21, Simo Sorce wrote:
Hi Matthew, you say: "We're missing people", and I think, "who?". And who are you going to miss if you move to discourse?
I will be candid, I tried to use forums since the old phpBB times, it never works for me. I have no time to go roaming over forums except if a search engine brings me there.
The mailing list make messages land in my client, on which I am very efficient, therefore I can check all messages once a day, and respond if I find a worthy topic.
Unless this discourse has some great mail bridge (it doesn't) or maybe an rss feed (I do not use those at work, but I guess I could ?) So that I can skim messages on my terms, I think I (and those like me) will be the next "missing people".
Ditto.
I am in the same boat, FWIW ...
I actually quite like Discourse - for a forum software - from experience related to various freetime activities.
However, Discourse replacing mailing lists WILL be the end of habitually skimming through everything that goes on devel (and a whole bunch of other lists) to spot issues that might be of my concern. The result will be in me being considerably less aware of what goes around in Fedora, rpm related or not.
Dan
On Fri, Apr 21, 2023 at 10:31:04AM +0300, Panu Matilainen wrote:
I actually quite like Discourse - for a forum software - from experience related to various freetime activities.
However, Discourse replacing mailing lists WILL be the end of habitually skimming through everything that goes on devel (and a whole bunch of other lists) to spot issues that might be of my concern. The result will be in me being considerably less aware of what goes around in Fedora, rpm related or not.
You can get notifications of every post in tags you want to follow by email, and presumably skim them in the same way. Maybe there's something we could do to make that better for you -- can you help me understand what doesn't work well?
We have another feature which might be of interest: we have enabled the "saved searches" plugin. You provide search queries, and whenever there is a new patch, you get a notification. (You can decide if you want these immediately, hourly, daily, or weekly.)
https://discussion.fedoraproject.org/t/new-site-feature-saved-searches-get-n...
I know Matthew mentioned long replies to long posts as a problem, so I'll do what on discourse would be an emoticon to sum a complex discussion:
+1 Simo :)
Cheers, Mario
On Fri, Apr 21, 2023 at 1:22 AM Simo Sorce simo@redhat.com wrote:
Hi Matthew, you say: "We're missing people", and I think, "who?". And who are you going to miss if you move to discourse?
I will be candid, I tried to use forums since the old phpBB times, it never works for me. I have no time to go roaming over forums except if a search engine brings me there.
The mailing list make messages land in my client, on which I am very efficient, therefore I can check all messages once a day, and respond if I find a worthy topic.
Unless this discourse has some great mail bridge (it doesn't) or maybe an rss feed (I do not use those at work, but I guess I could ?) So that I can skim messages on my terms, I think I (and those like me) will be the next "missing people".
Btw I could make exactly the same quote about any forum that Major made for Mailing lists, messy discussions are messy and a forum does not make them easier to follow by any means (perhaps except for those that chose inferior email readers).
All that said, why waste time with this discussion?
Your own post communicates to me (whether you intended it or not) that in the end the thread that will be generated by this post won't matter, because this is just a courtesy post and you already think that the opinion of the "minority of self selected mailing list lovers and dinosaurs" does not matter much.
On Thu, 2023-04-20 at 17:20 -0400, Matthew Miller wrote:
It’s time to transform the Fedora devel list into something new
-- Simo Sorce RHEL Crypto Team Red Hat, Inc
devel mailing list -- devel@lists.fedoraproject.org To unsubscribe send an email to devel-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
Hi,
On Thu, Apr 20, 2023 at 07:21:54PM -0400, Simo Sorce wrote:
The mailing list make messages land in my client, on which I am very efficient, therefore I can check all messages once a day, and respond if I find a worthy topic.
Unless this discourse has some great mail bridge (it doesn't) or maybe an rss feed (I do not use those at work, but I guess I could ?) So that I can skim messages on my terms, I think I (and those like me) will be the next "missing people".
I think I would fall in the same category. The fedora-devel mailinglist keeps me connected to the project even though I rarely reply to messages.
Your own post communicates to me (whether you intended it or not) that in the end the thread that will be generated by this post won't matter, because this is just a courtesy post and you already think that the opinion of the "minority of self selected mailing list lovers and dinosaurs" does not matter much.
Agreed. It was a very long post with several important and interesting observations about communication in a large groups of people. But it felt like it wasn't meant as actually discussing how we can communicate better, but to work towards a conclusion that a specific forum technology should be adopted and get rid of people who use the mailinglist to participate in the project.
Cheers,
Mark
Like many of you, I have built up a lot of workflow around my email. I'm on several dozen Fedora mailing lists, so I heavily filter my inbox. In fact, that's one of the areas where I'm least enthusiastic about a large-scale move to Fedora Discussion. Our tag-based setup makes it very difficult for Gmail (as an email client specifically) to filter alerts the way I want.
On the other hand, as someone who often needs to cross-post announcements and the like, the experience for that on Discussion is so much better. And there are a lot of other quality of life improvements, some big and some small, but they add up:
* The ability to move threads from "contributor" to "developer" areas and vice versa * Splitting off tangents to a new thread * Editing posts (with visible history, of course) * In-post polls * In-line date/time tags that can render the time in the user's zone
Discussion isn't perfect, but it's better on the whole than I thought it would be.
On Fri, Apr 21, 2023 at 1:22 AM Simo Sorce simo@redhat.com wrote:
Hi Matthew, you say: "We're missing people", and I think, "who?". And who are you going to miss if you move to discourse?
I will be candid, I tried to use forums since the old phpBB times, it never works for me. I have no time to go roaming over forums except if a search engine brings me there.
The mailing list make messages land in my client, on which I am very efficient, therefore I can check all messages once a day, and respond if I find a worthy topic.
Unless this discourse has some great mail bridge (it doesn't) or maybe an rss feed (I do not use those at work, but I guess I could ?) So that I can skim messages on my terms, I think I (and those like me) will be the next "missing people".
Btw I could make exactly the same quote about any forum that Major made for Mailing lists, messy discussions are messy and a forum does not make them easier to follow by any means (perhaps except for those that chose inferior email readers).
All that said, why waste time with this discussion?
Your own post communicates to me (whether you intended it or not) that in the end the thread that will be generated by this post won't matter, because this is just a courtesy post and you already think that the opinion of the "minority of self selected mailing list lovers and dinosaurs" does not matter much.
I have to say, even at the risk of making me sound like "old dinosaur man yells at cloud", I agree here.
The way I interact with the mailing list is very efficient - new messages land in my email client, they are clearly labeled as "unread", and archiving read messages (or discussion threads that I'm not interested in) creates a searchable archive of everything that happened "as a byproduct". Interacting with discourse works completely differently - messages don't "come to me", I need to *actively seek them out*, it's hard to tell what's new, the interface isn't entirely obvious (maybe this would be better if I used discourse more regularly, but I really don't want to do that, unless I *absolutely need to*).
Especially the "messages come to me" vs. "I need to go looking" aspect would likely result in me both 1. not being as informed about what's going on in Fedora, and 2. interacting much less with discussions than I do now.
Fabio
On Thu, Apr 20, 2023 at 07:21:54PM -0400, Simo Sorce wrote: ...snip...
Unless this discourse has some great mail bridge (it doesn't) or maybe an rss feed (I do not use those at work, but I guess I could ?) So that I can skim messages on my terms, I think I (and those like me) will be the next "missing people".
So I've been using the email bridge for a while (I think since we set it up) and it's got it's issues for sure, but I am not sure if it's as bad as folks fear.
https://discussion.fedoraproject.org/t/guide-to-interacting-with-this-site-b... has general info.
I had just been dumping it into one mailbox, but today I poked at getting it sorted better. For those of you ancient dinosaurs like myself still using procmail (written in 1990!), the following hacky recipe works for me:
:0 * ^List-Id: .*</[-a-z0-9]*.discussion.fedoraproject.org> { NAME=`echo "${MATCH}" | sed 's/.discussion.fedoraproject.org>$//'`
:0 $HOME/Maildir/.fedora.discussion.$NAME/ }
This gets posts flowing into folders by list-id, so:
.fedora.discussion.Ask-Fedora_Ask-in-English/ .fedora.discussion.Project-Discussion/
And of course you can filter more with the actual tags from there if you like. Posts should work fine as they have a reply-to hash with the topic/post and who the email was sent to.
I have been pondering if we could perhaps setup a public-inbox read-only mirror of the posts to discussion. ( https://public-inbox.org/README.html ). It would take a bit of work, as I think we would need to make a non priv user, subscribe to everything, then mangle the emails as they come to not have the reply-to or anything else thats specific to the user. However, that could be a solution to longer term archiving of things, another way for casual people to read things and also allow a nntp frontend for the crazy nntp folks. ;) public-inbox is plain text only, so no images/html there. If there's enough interest in this I would be happy to work with folks who want to set this up.
For rss feeds, you can in general add '.rss' to any url on the forum.
ie, https://discussion.fedoraproject.org/tag/guide.rss will give you a rss feed of all the things tagged '#guide'.
Also, there's a 'latest posts' feed at: https://discussion.fedoraproject.org/posts.rss This should be all posts as they come
and a latest topics at: https://discussion.fedoraproject.org/latest.rss which is just the topics as they come (ie, the new/initial post in each thread).
I've also been using the rss feeds for a long while and they seem fine and reliable.
As a side note, I use RSS for tons of things. My setup is to run miniflux ( https://miniflux.app/ ) on my main server at home, then I use newsflash on my laptop or an android miniflut on android to read feeds. YMMV.
kevin
On Sat, Apr 22, 2023 at 03:04:26PM -0700, Kevin Fenzi wrote:
So I've been using the email bridge for a while (I think since we set it up) and it's got it's issues for sure, but I am not sure if it's as bad as folks fear.
https://discussion.fedoraproject.org/t/guide-to-interacting-with-this-site-b... has general info.
I had just been dumping it into one mailbox, but today I poked at getting it sorted better. For those of you ancient dinosaurs like myself still using procmail (written in 1990!), the following hacky recipe works for me:
:0
- ^List-Id: .*</[-a-z0-9]*.discussion.fedoraproject.org>
{ NAME=`echo "${MATCH}" | sed 's/.discussion.fedoraproject.org>$//'`
:0 $HOME/Maildir/.fedora.discussion.$NAME/}
This gets posts flowing into folders by list-id, so:
.fedora.discussion.Ask-Fedora_Ask-in-English/ .fedora.discussion.Project-Discussion/
And of course you can filter more with the actual tags from there if you like. Posts should work fine as they have a reply-to hash with the topic/post and who the email was sent to.
One bug I hit setting this up this weekend... some (but not all) of the categories appear to have newlines in the List-ID: header. ;(
ie,
List-ID: Fedora Discussion | Ask Fedora Ask in English <ask-in-english.ask-fedora.discussion.fedoraproject.org>
Hopefully this is a bug that can be fixed...
So the above won't really work as the MATCH doesn't include the second line with the actual listid in it, only the description. ;(
kevin
Kevin Fenzi wrote:
One bug I hit setting this up this weekend... some (but not all) of the categories appear to have newlines in the List-ID: header. ;(
ie,
List-ID: Fedora Discussion | Ask Fedora Ask in English <ask-in-english.ask-fedora.discussion.fedoraproject.org>
Hopefully this is a bug that can be fixed...
So the above won't really work as the MATCH doesn't include the second line with the actual listid in it, only the description. ;(
I think you'll have to blame Procmail for that. What you show is called folding in RFC 5322. It's valid syntax if the continuation line begins with whitespace, which it looks like it does. Folding is even recommended for lines longer than 78 characters. Programs that parse email are supposed to unfold folded lines.
The complexities of text-based protocols provide for so much fun!
Björn Persson
On Sun, Apr 23, 2023 at 11:21:19PM +0200, Björn Persson wrote:
Kevin Fenzi wrote:
One bug I hit setting this up this weekend... some (but not all) of the categories appear to have newlines in the List-ID: header. ;(
ie,
List-ID: Fedora Discussion | Ask Fedora Ask in English <ask-in-english.ask-fedora.discussion.fedoraproject.org>
Hopefully this is a bug that can be fixed...
So the above won't really work as the MATCH doesn't include the second line with the actual listid in it, only the description. ;(
I think you'll have to blame Procmail for that. What you show is called folding in RFC 5322. It's valid syntax if the continuation line begins with whitespace, which it looks like it does. Folding is even recommended for lines longer than 78 characters. Programs that parse email are supposed to unfold folded lines.
The complexities of text-based protocols provide for so much fun!
Indeed. Actually procmail matches on it fine, but it doesn't include the part after the newline in $MATCH, which makes it harder to parse out the second part if you want to use that to filter into a maildir with that name. :(
I can work around it, it just seems... unexpected and I wanted to mention it in case others were going down that path.
kevin
On Sat, Apr 22, 2023 at 03:04:26PM -0700, Kevin Fenzi wrote:
I have been pondering if we could perhaps setup a public-inbox read-only mirror of the posts to discussion. ( https://public-inbox.org/README.html ). It would take a bit of work, as I think we would need to make a non priv user, subscribe to everything, then mangle the emails as they come to not have the reply-to or anything else thats specific to the user. However, that could be a solution to longer term archiving of things, another way for casual people to read things and also allow a nntp frontend for the crazy nntp folks. ;) public-inbox is plain text only, so no images/html there. If there's enough interest in this I would be happy to work with folks who want to set this up.
Rather than email subscription, this could be via a webhook. That payload includes both the html and "uncooked" message, a bunch of metadata. And it can trigger on post, edit, metadata change, and other things. That's probably a better interface than email mangling.
On Tue, Apr 25, 2023 at 03:29:34PM -0400, Matthew Miller wrote:
On Sat, Apr 22, 2023 at 03:04:26PM -0700, Kevin Fenzi wrote:
I have been pondering if we could perhaps setup a public-inbox read-only mirror of the posts to discussion. ( https://public-inbox.org/README.html ). It would take a bit of work, as I think we would need to make a non priv user, subscribe to everything, then mangle the emails as they come to not have the reply-to or anything else thats specific to the user. However, that could be a solution to longer term archiving of things, another way for casual people to read things and also allow a nntp frontend for the crazy nntp folks. ;) public-inbox is plain text only, so no images/html there. If there's enough interest in this I would be happy to work with folks who want to set this up.
Rather than email subscription, this could be via a webhook. That payload includes both the html and "uncooked" message, a bunch of metadata. And it can trigger on post, edit, metadata change, and other things. That's probably a better interface than email mangling.
yeah, that would be great if it can use a webhook.
kevin
On Thu, Apr 20, 2023 at 05:20:37PM -0400, Matthew Miller wrote:
We’re missing people
A Mastodon post from long-time Fedora contributor Major Hayden got me thinking:
How do people make so much time available for mailing list discourse?
Once I ensure my team has the technical guidance they need and I work through the tasks of work that I owe other people, I take a look at the mailing list and say: "Oh my gosh, what the heck happened here?" Then the discussion goes further off the rails while I'm typing out a reply and my reply is no longer relevant.
I know many Fedora folks, old-school and new, for whom devel list is just too much.
I think that applies to the vast majority of people on my team. They are doing work that ends up in Fedora, they use Fedora, and they will sometimes needs to make changes in Fedora packages, but they'll rarely / never read the mailing lists. Essentially, unless a very large amount of your daily work is Fedora oriented, the devel list is not feasible to follow. It is not suited to infrequent/sporadic Fedora contributors.
Myself I'll try to keep an eye out for interesting $SUBJECTs and read those, or peek into exploding threads to see what the fuss is about, but the rest I'll just ignore as it is too much. I'll try to relay the important nuggets of info to others in my team, but that's a limited mitigation.
When you say "We're missing people" there's actually another factor that you've not mentioned....spam, or more specifically anti-spam countermeasures.
There are a handful of very regular Fedora contributors, for whom about 50% of messages they send get reliably classified as spam by my employer's mail service because of something it dislikes about their mail server/domain's reputation. While I can allow-list their addrs for myself, each subscriber has to repeat the same allow-listing and I expect many won't bother.
IOW, we have people who think they are contributing to Fedora discussions, but in fact their mails are getting effectively silently discarded. This is bad if we want inclusive discussions.
In other mailing lists where I am the admin I get to see even worse problems where mails get unconditionally discarded at time of delivery due to disagreements between mail servers about the correct way to implement DMARC and DKIM. As a result certain senders are entirely prevented from collaborating via the mailing list in the worst possible way. They successfull send the mail, it is accepted by the list server, added to the online archives, but when the list server delivers to subscribers the message bounces back. Unless the mailing list admin watches non-delivery bounces no one will know this is happening. Certain senders will simply not be part of the discussions.
Having battled email/spam problems wrt mailing lists for years now, I can only conclude that email is not viable as a reliable & inclusive communications tool in the modern world.
We’re scattered in actual practice
Many groups have actually moved away from lists to using tickets for team conversations — both those non-engineering functions and development. Design Team has a mailing list, but mostly for announcements: the work happens in tickets. Workstation largely uses their Pagure tracker. And CoreOS conversations happen almost entirely in tickets on Github.
Tickets are made for tracking specific, actionable tasks, and that kind of tracking is part of why teams use them over mailing lists — but Fedora teams use them for open development conversations too. I think that’s largely a symptom of mailing lists not being enough for what we need. The trackers have media support, editing for typos or updates, reactions for simple agreement, tagging people, and granular subscriptions. They are effectively “off-label use” mini-forums that teams can quietly move to using without the sort of conversation I expect this message to generate.
Airplane diagram, survivorship bias
The set of remaining regular participants on this list is naturally biased towards those for whom it is working just fine. But, core Fedora development discussion can’t be limited to that ever-shrinking group. Consider who isn’t here. The problems are real, and the trend isn’t in a good direction.
Or stockholm syndrome. As a long term email users I'm familiar with it and learn to put up with all its flaws and figured out ways to mitigate the limitations of email.
If you speak to any long term email users in high traffic OSS communities you'll find many of them have created very elaborate mail filters, scripts and tools to create workflows for handling the high volume of their inbox. Newer contributors won't have this, so when you ask them to turn on the fire hose by subscribing to Fedora devel it is natural they will be reluctant.
Concrete proposal
I’m not suggesting we shut down devel list next week. And I think we’ll have some mailing lists for quite a long time. But, I think it’s time to start moving some specific things, with the eventual goal of closing every mailing list we can.
First, I’d like to move the Changes discussion. They will still be posted to devel-announce, but responses directed to Project Discussion in a new #changes tag. Ben tells me that this is a FESCo decision, which seems reasonable.
Early you mentioned that lots of teams have moved discussions into tickets in Pagure/GitHub/etc. I think that is an entirely natural thing to do for task oriented discussions.
I think that Change proposals are precisely that - a task oriented discussion, and thus would naturally fit into issue tracker model. You then have the associated tools for tracking, tagging the changes, linking between related tickets, and more, all in one place. The way we use wiki categories to tag Change proposal pages, and then have to have discussions somewhere else entirely, is effectively reinventing a poor-mans' issue tracker. Just use a real issue tracker here and stop splitting the process around multiple tools.
Next steps
I know this is a big change. I’ve been thinking of writing this message for a long time. I’d really like to convince everyone that it’s the right thing — or at least, an acceptable one.
I don't disagree, and as you illustrate we would not be going into uncharted territory here. Many big OSS projects have gone the same way for similar reasons. Fedora is about innovation, being on the cutting edge, etc, and I think that includes being willing to adapt to changes in the way we work as a community too, not be left behind using obsolete processes and tools that are increasingly abandoned by the rest of the OSS community.
This doesn't mean we're saying all aspects of email are bad. There are many things to like about email, and reducing/eliminating its use will certainly make some things worse. Your proposal is effectivey saying that when considered in aggregate, switching to newer tools like discourse and/or issue trackers (as appropriate for the particular scenario) will still be a net win, despite some of the downsides.
Much as people invaribly hate change, they are also very adaptable. Having seen these kind of transitions before, my experiance is that the downsides never turn out to be as bad as imagined because once people get past the grief & anger about the change, they ultimately figure out alternative ways to work effectively in many cases.
With regards, Daniel
Hi,
On 4/21/23 10:47, Daniel P. Berrangé wrote:
I agree with most of your points, but I wanted to comment on the Change process.
(...skipped a lot...)
First, I’d like to move the Changes discussion. They will still be posted to devel-announce, but responses directed to Project Discussion in a new #changes tag. Ben tells me that this is a FESCo decision, which seems reasonable.
Early you mentioned that lots of teams have moved discussions into tickets in Pagure/GitHub/etc. I think that is an entirely natural thing to do for task oriented discussions.
I think that Change proposals are precisely that - a task oriented discussion, and thus would naturally fit into issue tracker model. You then have the associated tools for tracking, tagging the changes, linking between related tickets, and more, all in one place. The way we use wiki categories to tag Change proposal pages, and then have to have discussions somewhere else entirely, is effectively reinventing a poor-mans' issue tracker. Just use a real issue tracker here and stop splitting the process around multiple tools.
I think the issue with the Change process is that while the change itself is a task, its discussion has a property of generating sub-threads and sidetracks going in all directions, which shape the rest of the fedora-devel discourse.
So while from the execution point of view having it in the issue tracker makes a lot of sense, from the communication point of view Change discussion is a seed, which when planted on a forum can create a whole forest.
The hope is that forum tooling can be used to split the subthreads and to let them live on their own. So instead of strict moderation of the conversation in the issue tracker, which would be required to keep it on point, we will allow conversations to spread out but in a way that you have a possibility to choose your track.
P.S. It is funny how even though I consider myself a "seasoned" Fedora Contributor, every time I send a mail to a mailing list, I still go to a Hyper-Kitty interface to verify that I did everything correctly.
On 4/20/23 23:20, Matthew Miller wrote:
It’s time to transform the Fedora devel list into something new
My first post on this list was over 19 years ago. (It was about Bugzilla. I was a fan!) Ever since those early days, devel list has been the heart and center of Fedora activity. Now, I hope to convince all of us here that it’s time for something different.
As it is, devel list is too much for many people to follow — people we’d like to have around. It covers many different things at once, yet also drives us towards more scattered communications. Our infamous mega-threads are not really effective for getting to community consensus, and tend to bring out the worst in us.
Passionate people generate passionate discussion.
The only thing you will gain by a forum is that at the point the message will not be deemed appropriate, it will probably be deleted or "beatified" by the mod team. The passion from our human nature will not go away with a platform change.
I propose that we transition devel list, and eventually most of our mailing lists, to Fedora Discussion (our Discourse-powered forum).
I know this is a big change, but, hear me out…
We’re missing people
A Mastodon post from long-time Fedora contributor Major Hayden got me thinking:
How do people make so much time available for mailing list discourse?
Once I ensure my team has the technical guidance they need and I work through the tasks of work that I owe other people, I take a look at the mailing list and say: "Oh my gosh, what the heck happened here?" Then the discussion goes further off the rails while I'm typing out a reply and my reply is no longer relevant.
I know many Fedora folks, old-school and new, for whom devel list is just too much. Some of it is the sheer volume, but this “off the rails” tendency is real — threads drift, get into back-and-forth debates about particular details, etc.
Moving a discussion does not reduce the volume. From what I've heard from engineers around me, the volume of devel is hard to keep up with. That, to me, does not sound like a platform problem, but a scale problem.
And… some people aren’t here because — in contrast with our “Friends” foundation, it isn’t always a nice place to be (and mailing lists don’t provide many tools for moderation, except the big hammer of outright bans).
Ben Cotton recently did some basic analysis on devel list traffic over time, and there’s a clear trend: fewer people are participating, even though the number of different threads goes up. I don’t think this is because of any decline in Fedora contributors overall — I think it’s that conversations are happening elsewhere.
Big threads are … bad, actually
When we have something to talk about, it tends to explode into a big thread. The thing in January with FESCo’s frame pointers decision (https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org/...) is a good example of things going badly.
Most of the conversation was under the subject “Schedule for Tuesday's FESCo Meeting (2023-01-03)”, because everything started as a reply to that. That’s pretty easy to overlook. It’s possible for replies to change the subject when replying, but that can’t be done retroactively, and then isn’t consistent (and it breaks threading in Gmail, too).
Then, things got rather hostile, making it hard to have a reasonable conversation about the issues (both technical and procedural). And then, things went in circles without adding anything new.
This could have all gone a lot better.
And that’s just one example. Take a look back at any mega-thread, and you’ll find similar — and worse. When things get heated, the only way to intervene is by adding more. There are often long subthreads of two people going back and forth on tangents. Then, other conversation branches duplicate that, or refer across. Classical email tools don’t actually handle this kind of thing very well at all. In my experience, it only really works if you keep up with the conversation in almost real time, which has its own problems even when that’s possible.
We’re scattered in actual practice
Devel list may be the center, but we have _hundreds_ of Fedora mailing lists. A dozen or so are reasonably active (Test, Legal, ARM…) but most are inactive or dead. Some are just meeting reminders over and over — for meetings that aren’t even active anymore. It’s easy to make but hard to _unmake_ a mailing list.
For lists that are active, the split is confusing — when should something be on the packaging list rather than devel? What happens when something is related to both Cloud and Server, or Workstation and KDE? One can post to both lists, but if someone replies and isn’t subscribed to both, the conversation gets split.
With “devel” as the main list, conversations about marketing, design, events, and so on don’t really have a central place. (The Mindshare list never really caught on.) That makes these important activities feel even more disconnected and secondary in status — and they shouldn’t be.
Many groups have actually moved away from lists to using tickets for team conversations — both those non-engineering functions and development. Design Team has a mailing list, but mostly for announcements: the work happens in tickets. Workstation largely uses their Pagure tracker. And CoreOS conversations happen almost entirely in tickets on Github.
Tickets are made for tracking specific, actionable tasks, and that kind of tracking is part of why teams use them over mailing lists — but Fedora teams use them for open development conversations too. I think that’s largely a symptom of mailing lists not being enough for what we need. The trackers have media support, editing for typos or updates, reactions for simple agreement, tagging people, and granular subscriptions. They are effectively “off-label use” mini-forums that teams can quietly move to using without the sort of conversation I expect this message to generate.
It's hard to not read this as a search for catch-all solutions.
Personally, as a packager, if I want Fedora-wide visibility for a change, I will create a change proposal. That change proposal was created as a reaction, generally, to a ticket (e.g. Bug FooBar makes Baz hard to use) that has wider consequences.
A discussion to a technical change, for me, will forever be in a ticket. No matter the "wider discussion platform" projects will always have bug trackers where one can create a ticket.
Airplane diagram, survivorship bias
Since I’m posting this to devel list, I do expect a lot of push-back. Maybe I’ll be surprised and more of y’all are already with me on this and just waiting for something to happen. But overall I expect a tough crowd.
You’ve probably heard the story about bullet holes in airplanes returning from missions and the accompanying diagram — find it athere https://en.wikipedia.org/wiki/Survivorship_bias if that’s new to you.
The set of remaining regular participants on this list is naturally biased towards those for whom it is working just fine. But, core Fedora development discussion can’t be limited to that ever-shrinking group. Consider who isn’t here. The problems are real, and the trend isn’t in a good direction.
But, is it shrinking due to a platform, or something other?
It makes me want to try discourse out, not saying I'll stick around, I am, luckily, not paid to read forums with no threading. IMO, a stream of posts with mentions of previous posts is not threading. Threading begins and ends on new topic posts AFAICT on discourse.
But I'd be happier if there was some tangible metric how to measure if we got more *related to the topic* engagement. I would hate to see 20 "+1" posts from "random" users counted towards "it is better now".
Devel List is too many different things!
We use it for Change discussion (resulting in those not-actually-so-great big threads).
We use it for introductions and onboarding — we’re usually pretty good at that, actually (but it adds to the overall load of following the list). We’re not very consistent, though.
We discuss packaging: guidelines, help on different topics, coordination on specific work. There’s unclear overlap with the packaging list, though, which is a bit confusing.
We talk about higher-level Fedora OS development topics that don’t fit anywhere else. For example, this on installer environment size: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org/... Sometimes, these are FESCo topics (which FESCo heroically struggles to keep here rather than on Pagure tickets…).
Sometimes, people talk about a particular interest in the hopes of forming a Special Interest Group. Often that does result in a new SIG launching — sometimes with a brand-new mailing list which then ends up getting only a few posts ( and a whole lot of spam years later). But this really suffers from the survivorship problem: many people who might be interested will never notice.
We have release process announcements — mostly to devel-announce, but then occasionally replies and discussion here. Blocker bugs and other test-list and QA topics are cross-posted, as part of the release cycle.
And then there are a lot of robo-messages: reports, reminders, etc. These are really valuable to a few people, but add a lot more to wade through for everyone else trying to keep up.
There are certainly others, as well.
In short… there probably shouldn’t just be one thing. But, the cross-posting problem makes it hard to split up as a mailing list.
Enter Discourse
If you’ve talked with me about anything related to any of this in the past ten years, you probably already know that I like Discourse. It’s good software made by cool people. And, it’s entirely open source, with a SaaS business model but with real, usable releases. (No open core, no “open source theoretically but good luck”.) And, we have it in production in Fedora already, athttps://discussion.fedoraproject.org, with categories for announcements, user help, project discussion, and social conversation — as well as special categories for dedicated workflows.
In Project Discussion, each different Fedora team can have its own tag, and you can subscribe to those that you’re interested in. Cross-posting is easy: tag a post with multiple teams.
I'd be interested in having a kind of "crossroad sign", to direct me towards tags what I would care about from a packager perspective. Not happy about this change, but it would make my experience a bit better...
Topics can be renamed, if needed, or split out into side-conversations. The long tangents from these conversations can actually be interesting on their own without distracting from the main topic. Moderation tools allow us to handle posts that are outside of expected Fedora contributor behavior, with varying levels of action as appropriate.
You can use markdown formatting, including images (with easy addition of alt text for people for whom images are a barrier). You can edit your posts to fix typos or correct mistakes. There are polls and lots of other useful features.
And, you can interact with it all by email. That interface isn’t perfect, but it’s way better than any other forum software I’ve seen. (I’ve written a guide:https://discussion.fedoraproject.org/t/25960) At the same time, your individual email address is hidden, so it won’t be a spam magnet (or a target for off-list harassment, a problem we unfortunately have with devel list).
That said, it is web-first software. (Or mobile, if that’s your thing.) That requires some adjustment, I know. I hope opening up a Fedora Discussion tab – or keeping one open — becomes an easy habit.
If I was a volunteer that's the thing I'd remember once in a blue moon that it even exists. But I guess that's just person to person :).
Not just Fedora
There’s a big trend towards Discourse in open source projects overall. Python and Gnome have both migrated entirely from their mailing lists. Ansible is working on it. Plus, there’s Rust, Kubernetes, Nextcloud, Flathub, Grafana, Home Assistant, KDE, and I’m sure many others.
That sounds non-ideal, glad I don't have to deal with those projects. If it works, great for them. Personally, a drive-by contribution with a question to a mailing list is much easier than having to sign up. I mean I do sign up occasionally, but it's with a kind of 10minute mail sign up generally.
Concrete proposal
I’m not suggesting we shut down devel list next week. And I think we’ll have some mailing lists for quite a long time. But, I think it’s time to start moving some specific things, with the eventual goal of closing every mailing list we can.
First, I’d like to move the Changes discussion. They will still be posted to devel-announce, but responses directed to Project Discussion in a new #changes tag. Ben tells me that this is a FESCo decision, which seems reasonable.
Second, I think other FESCo-related conversations should move. I hope this will reduce the urge to have back-and-forth exchanges in the tickets. For the Fedora Council, I set up a bot which automatically creates a discussion topic when a ticket is filed, leaving the ticket just for votes and recording of outcome. FESCo could use something similar.
Third, I’ll add a tag for general Fedora OS development conversation. Maybe “#devel”, but if someone has more narrow suggestions, I’ll take them.
Fourth, I’d like to update our documentation, process, and expectations for newcomers — say hello on Discussion (and Fedora Chat, if you like) rather than a mailing list. (I’d like to close the Fedora Join list at this point.)
Fifth, all packaging-related discussions (including the separate packaging mailing list). We already have a #package-maintainers tag with some existing discussion.
Sixth, automated posts, as much as we can. These should go to dedicated Workflow categories, where people who want can watch them but where they won’t overwhelm human interactions. People who want can watch them, and it’s easy to quote-reply into a new linked topic in the Project Discussions category.
And finally… shut down the devel list itself. Perhaps at the end of 2023?
We should also shut down all of the little lists that haven’t had anything but spam in the last two years. We could maybe do that sooner. We should stop creating new lists now — we can create new Discussion tags instead.
I expect the announcement lists to stay for the foreseeable future (although we might feed them from Discourse rather than the other way around). Other lists which are patches, commit messages, and other automations might stick around for a while — but really might be better served by a log aggregation and analysis system or something else.
Other teams who want to keep mailing lists can, but I’d like to move those too, and eventually I think we’ll want to shut them down too — or perhaps convert them to announcement lists.
Next steps
I know this is a big change. I’ve been thinking of writing this message for a long time. I’d really like to convince everyone that it’s the right thing — or at least, an acceptable one.
I'll be convinced once I see it all.
What about specific decisions related to my proposal? For each:
Because altering the Changes process is a FESCo decision, I’ll file a ticket about that shortly.
FESCo moving their own other conversations is, of course, also a FESCo decision.
Assuming the first moves forward, I will create the general #devel tag (or other name we come up with) when I create the #change-proposal tag.
Moving the packaging list is a Packaging Committee decision.
Automated posts can be moved at any time. I can work with the people who own the generation of those reports to figure out a good answer for each.
The outcome for other team lists is up to each team.
And, I think shutting down devel overall is ultimately a Fedora Council decision.
For right now, though: let’s discuss — on the list!
As a person in my early 20s, I hate how everything is becoming web centric and no one can convince me to feel otherwise. While I am hearing from varying people around me, how it must be bad using email, it provides client-side filtering unparalleled by any platform that I used in the past.
I enjoyed Fedora being on mailing lists, nothing ever came close to the threading of emails. I was not getting lost in threads of conversation while still being under the umbrella topic, no need to open who knows how many links to read all tangents.
Regards, Jarek Prokop Software Engineer – Core Services Apps, Ruby RH
Hi all,
On 21/04/2023 10:50, Jarek Prokop wrote:
As a person in my early 20s, I hate how everything is becoming web centric and no one can convince me to feel otherwise.
hm... I thought this was kind of a generation conflict. Glad to be proven wrong.
I enjoyed Fedora being on mailing lists, nothing ever came close to the threading of emails.
Actually, Zulip [1] does. It's an interesting tool, basically something in between traditional forums and email threading.
--alec
On Fri, Apr 21, 2023 at 10:50:42AM +0200, Jarek Prokop wrote:
also drives us towards more scattered communications. Our infamous mega-threads are not really effective for getting to community consensus, and tend to bring out the worst in us.
Passionate people generate passionate discussion.
The only thing you will gain by a forum is that at the point the message will not be deemed appropriate, it will probably be deleted or "beatified" by the mod team. The passion from our human nature will not go away with a platform change.
That's true -- and I'm not looking to get rid of passion, or silence opinions. But when something is _really_ out of line (often written in the heat of the moment), it's better to have options to ... as you say, beautify* the conversation. That makes it better for other people participating, and better for the person who has a chance to make their point in a more constructive way.
* also, to fix typos :)
[snip]
A discussion to a technical change, for me, will forever be in a ticket. No matter the "wider discussion platform" projects will always have bug trackers where one can create a ticket.
Of course. That's not what I'm talking about. Consider for example this: https://pagure.io/fesco/issue/2817. That's not about the technical decision itself -- it's an branch of the conversation that should have been here.
biased towards those for whom it is working just fine. But, core Fedora development discussion can’t be limited to that ever-shrinking group. Consider who isn’t here. The problems are real, and the trend isn’t in a good direction.
But, is it shrinking due to a platform, or something other?
I don't think Fedora contribution and activity overall are shrinking. And I'm quite convinced that the platform is part of it.
It makes me want to try discourse out, not saying I'll stick around,
I'm glad to hear that.
I am, luckily, not paid to read forums with no threading. IMO, a stream of posts with mentions of previous posts is not threading. Threading begins and ends on new topic posts AFAICT on discourse.
It's not presented as a tree, but there _are_ threads of replies. If you see something like "2 replies" under a particular post, you can click that and the view will be restricted to just those replies, which you can then follow further.
Example: https://discussion.fedoraproject.org/t/future-of-encryption-in-fedora-deskto...
But also, yes — when something really diverges in Discourse, it should be a new topic. A moderator can move things after the fact (like I did with https://discussion.fedoraproject.org/t/getting-systemd-homed-working-properl...) but even better, when replying, you can create a linked topic. See https://discussion.fedoraproject.org/t/site-tip-create-linked-topics-for-dee...
But I'd be happier if there was some tangible metric how to measure if we got more *related to the topic* engagement. I would hate to see 20 "+1" posts from "random" users counted towards "it is better now".
That's reasonable. Do you have suggestions for a good metric?
In Project Discussion, each different Fedora team can have its own tag, and you can subscribe to those that you’re interested in. Cross-posting is easy: tag a post with multiple teams.
I'd be interested in having a kind of "crossroad sign", to direct me towards tags what I would care about from a packager perspective. Not happy about this change, but it would make my experience a bit better...
There's a big _index_ at https://discussion.fedoraproject.org/tags, but that's probably bit much (while at the same time not containing enough description). What would this "sign" ideally look like to you?
That said, it is web-first software. (Or mobile, if that’s your thing.) That requires some adjustment, I know. I hope opening up a Fedora Discussion tab – or keeping one open — becomes an easy habit.
If I was a volunteer that's the thing I'd remember once in a blue moon that it even exists. But I guess that's just person to person :).
There _are_ email notifications, and you can interact by replying to them. (You can even +1 or <3.)
There is also a "digest" mail sent automatically if you're not active, showing active topics possibly of interest, which can serve as a more-frequent-than-blue-moon reminder. (You can turn this off, of course.)
As a person in my early 20s, I hate how everything is becoming web centric and no one can convince me to feel otherwise. While I am hearing from varying people around me, how it must be bad using email, it provides client-side filtering unparalleled by any platform that I used in the past.
It's fine, but it's no NNTP. That was really the best. :)
Do take a look at
https://discussion.fedoraproject.org/t/guide-to-interacting-with-this-site-b...
It's not perfect, but it's better than most other forum software's email interfaces.
I enjoyed Fedora being on mailing lists, nothing ever came close to the threading of emails. I was not getting lost in threads of conversation while still being under the umbrella topic, no need to open who knows how many links to read all tangents.
I appreciate your perspective, feedback, and willingness to try this out!
Once upon a time, Matthew Miller mattdm@fedoraproject.org said:
- also, to fix typos :)
So, I will say this is kind of a peeve of mine about server-based discussion systems (whether web or client like Slack/Discord): allowing people to edit messages, especially after people have replied to them, is a bad idea. Person 1 says "we should do XYZ", somebody replies "no XYZ is bad", and person 1 can go change their original message to say something completely different.
It kind of goes back to who "owns" (and I don't mean in the legal sense) the content. When the content is held on a server, the server owner has an editorial control that can be problematic.
Fixing typos sounds nice, but... just don't make typos, or proofread. :)
On Fri, Apr 21, 2023 at 12:07 PM Chris Adams linux@cmadams.net wrote:
So, I will say this is kind of a peeve of mine about server-based discussion systems (whether web or client like Slack/Discord): allowing people to edit messages, especially after people have replied to them, is a bad idea. Person 1 says "we should do XYZ", somebody replies "no XYZ is bad", and person 1 can go change their original message to say something completely different.
It kind of goes back to who "owns" (and I don't mean in the legal sense) the content. When the content is held on a server, the server owner has an editorial control that can be problematic.
Fixing typos sounds nice, but... just don't make typos, or proofread. :)
Chris Adams linux@cmadams.net
Usually when people edit messages there's a notation that it was edited. Some even note the number of times something has been edited, and by whom, including when mods/admin edit it. I forget what options discourse has for that. But also, when you are quoting someone on a forum, if they go back and change their message, i dont recall ever seeing a forum where it auto-updated the quote replies. So the person's original message will still show, however that does cause a bit of confusion when a new reader comes to the thread and sees a quote that doesn't match what they see in that person's post. There was a forum somewhere that the edit message would say "This post was edited by ${user} after replies by ${user2}, ${user3}, ${user4}" which helped eliminate that... but I believe that may have been something custom that the admin put in place. Also there's usually a time limit in place for edits... I've seen it as low as 30 seconds on some forums which negates some of this issue.
On 4/21/23 18:07, Chris Adams wrote:
Once upon a time, Matthew Millermattdm@fedoraproject.org said:
- also, to fix typos :)
So, I will say this is kind of a peeve of mine about server-based discussion systems (whether web or client like Slack/Discord): allowing people to edit messages, especially after people have replied to them, is a bad idea. Person 1 says "we should do XYZ", somebody replies "no XYZ is bad", and person 1 can go change their original message to say something completely different.
Hmm, interesting, seems like the platform is able to do a message based diff,
so if you edit your message, history is actually visible!
See for example this message from some of Matt's links I got open https://discussion.fedoraproject.org/t/future-of-encryption-in-fedora-deskto... There is a pencil with a number, click that, you can see a diff.
So I don't think that I have to be that concerned that someone will want to radically change their messages.
It kind of goes back to who "owns" (and I don't mean in the legal sense) the content. When the content is held on a server, the server owner has an editorial control that can be problematic.
Fixing typos sounds nice, but... just don't make typos, or proofread. :)
On Fri, Apr 21, 2023 at 11:07:16AM -0500, Chris Adams wrote:
Once upon a time, Matthew Miller mattdm@fedoraproject.org said:
- also, to fix typos :)
So, I will say this is kind of a peeve of mine about server-based discussion systems (whether web or client like Slack/Discord): allowing people to edit messages, especially after people have replied to them, is a bad idea. Person 1 says "we should do XYZ", somebody replies "no XYZ is bad", and person 1 can go change their original message to say something completely different.
Sure that is possible, but lets not assume nor optimize for malicious engagement. We should be optimizing for honest engagement because our contributors are decent people. The vast majority of the time when people go back to edit messages it will be to correct mistakes/typos, or syntax errors in formatting text, and so a clear benefit.
In the unlikely / rare event that someone did maliciously change the original message for some nefarious reason, other users can call them out, and in the worst case moderators could take action. I very much doubt that would happen enough to be a problem worth worrying about in a constructive community environment like Fedora.
With regards, Daniel
On Fri, Apr 21, 2023 at 11:07:16AM -0500, Chris Adams wrote:
Once upon a time, Matthew Miller mattdm@fedoraproject.org said:
- also, to fix typos :)
So, I will say this is kind of a peeve of mine about server-based discussion systems (whether web or client like Slack/Discord): allowing people to edit messages, especially after people have replied to them, is a bad idea. Person 1 says "we should do XYZ", somebody replies "no XYZ is bad", and person 1 can go change their original message to say something completely different.
Do you mean maliciously? In that case, it's a matter of asking people to not do that.
Or do you mean that it makes the history of the conversation confusing? The history _is_ there — edited posts are marked as such and you can see the changes.
I think in many cases it's fine for edited posts to reflect an updated understanding. If I post something and later realize I was confused, and so fix it, in a year no one will care about the initial mistake and it's more useful to have a "clean" post. (And any replies telling me I was wrong can be deleted, having served their purpose.)
Of course, that kind of "timeline edit" isn't appropriate for something that's a big decision or where the hashing-it-out _matters_. But it's very useful for posts like https://discussion.fedoraproject.org/t/fedora-strategy-2028-a-topic-index-fo...
Once upon a time, Matthew Miller mattdm@fedoraproject.org said:
On Fri, Apr 21, 2023 at 11:07:16AM -0500, Chris Adams wrote:
Once upon a time, Matthew Miller mattdm@fedoraproject.org said:
- also, to fix typos :)
So, I will say this is kind of a peeve of mine about server-based discussion systems (whether web or client like Slack/Discord): allowing people to edit messages, especially after people have replied to them, is a bad idea. Person 1 says "we should do XYZ", somebody replies "no XYZ is bad", and person 1 can go change their original message to say something completely different.
Do you mean maliciously? In that case, it's a matter of asking people to not do that.
I've seen it used that way, so that was my concern.
Or do you mean that it makes the history of the conversation confusing? The history _is_ there — edited posts are marked as such and you can see the changes.
But that definitely mitigates that concern. Most sites I've seen only show you the edited version, with no history available (at least to regular users).
On 4/21/23 17:42, Matthew Miller wrote:
On Fri, Apr 21, 2023 at 10:50:42AM +0200, Jarek Prokop wrote:
also drives us towards more scattered communications. Our infamous mega-threads are not really effective for getting to community consensus, and tend to bring out the worst in us.
Passionate people generate passionate discussion.
The only thing you will gain by a forum is that at the point the message will not be deemed appropriate, it will probably be deleted or "beatified" by the mod team. The passion from our human nature will not go away with a platform change.
That's true -- and I'm not looking to get rid of passion, or silence opinions. But when something is _really_ out of line (often written in the heat of the moment), it's better to have options to ... as you say, beautify* the conversation. That makes it better for other people participating, and better for the person who has a chance to make their point in a more constructive way.
- also, to fix typos :)
Oh, probably an important related feature I noticed after looking at Chris Adams' response, I had a small concern about people changing messages too radically, where the conversation will then lose meaning, the software actually supports history and colorful diffs.
[snip]
A discussion to a technical change, for me, will forever be in a ticket. No matter the "wider discussion platform" projects will always have bug trackers where one can create a ticket.
Of course. That's not what I'm talking about. Consider for example this: https://pagure.io/fesco/issue/2817. That's not about the technical decision itself -- it's an branch of the conversation that should have been here.
biased towards those for whom it is working just fine. But, core Fedora development discussion can’t be limited to that ever-shrinking group. Consider who isn’t here. The problems are real, and the trend isn’t in a good direction.
But, is it shrinking due to a platform, or something other?
I don't think Fedora contribution and activity overall are shrinking. And I'm quite convinced that the platform is part of it.
It makes me want to try discourse out, not saying I'll stick around,
I'm glad to hear that.
I am, luckily, not paid to read forums with no threading. IMO, a stream of posts with mentions of previous posts is not threading. Threading begins and ends on new topic posts AFAICT on discourse.
It's not presented as a tree, but there _are_ threads of replies.
Heh, sounds like a fun side project to try to transform it into a tree structure.
If you see something like "2 replies" under a particular post, you can click that and the view will be restricted to just those replies, which you can then follow further.
Example:https://discussion.fedoraproject.org/t/future-of-encryption-in-fedora-deskto...
Finally, noticed what it does, it made me a bit confused as the first response was the same as in the "global" flow of the topic, but the message under it changed. I think that it should be better visible that they are actually replies.
It seems to hide other replies and only show those that are part of the "thread". Do they accept RFEs? :) I think enhancing the visibility after I expand replies for the posts in the "thread" would be better.
But also, yes — when something really diverges in Discourse, it should be a new topic. A moderator can move things after the fact (like I did with https://discussion.fedoraproject.org/t/getting-systemd-homed-working-properl...) but even better, when replying, you can create a linked topic. See https://discussion.fedoraproject.org/t/site-tip-create-linked-topics-for-dee...
But I'd be happier if there was some tangible metric how to measure if we got more *related to the topic* engagement. I would hate to see 20 "+1" posts from "random" users counted towards "it is better now".
That's reasonable. Do you have suggestions for a good metric?
I'm afraid none that could be automated, but I am not one strong on metrics. I'll just throw out some ideas: 1. Number of unique contributors 2. How many unique posts these contributors interacted with 3. "quality" of the post. I think one could go by the length and verbosity of the post. E.g. "Yeah seems like a good change" is not as valuable as a deeper dive/analysis into a hypothetical problematic. (especially if we consider that the platform has +1 equivalents in reactions :)) 4. Number/frequency of interactions.
Maybe a combination of 1. and 4. would have value.
But we can worry about that a bit later than "right now".
In Project Discussion, each different Fedora team can have its own tag, and you can subscribe to those that you’re interested in. Cross-posting is easy: tag a post with multiple teams.
I'd be interested in having a kind of "crossroad sign", to direct me towards tags what I would care about from a packager perspective. Not happy about this change, but it would make my experience a bit better...
There's a big _index_ athttps://discussion.fedoraproject.org/tags, but that's probably bit much (while at the same time not containing enough description). What would this "sign" ideally look like to you?
Usually we put these into "might want to follow" in various resources.
Perhaps a relevant tag list in the Fedora documentation? There are various onboarding and "must read" level stuff where I think we point to Fedora devel mailing list (or equivalent for SIGs).
Basically a recommendation of "follow new proposals with tag #proposals, packaging questions with #packaging ...... tags on Fedora Discussion"
TBH, I enjoy simplicty, in time of writing, I was imagining something akin to a wiki page with a few lines on recommended tags.
That said, it is web-first software. (Or mobile, if that’s your thing.) That requires some adjustment, I know. I hope opening up a Fedora Discussion tab – or keeping one open — becomes an easy habit.
If I was a volunteer that's the thing I'd remember once in a blue moon that it even exists. But I guess that's just person to person :).
There _are_ email notifications, and you can interact by replying to them. (You can even +1 or <3.)
There is also a "digest" mail sent automatically if you're not active, showing active topics possibly of interest, which can serve as a more-frequent-than-blue-moon reminder. (You can turn this off, of course.)
That is good to know.
As a person in my early 20s, I hate how everything is becoming web centric and no one can convince me to feel otherwise. While I am hearing from varying people around me, how it must be bad using email, it provides client-side filtering unparalleled by any platform that I used in the past.
It's fine, but it's no NNTP. That was really the best. :)
Do take a look at
https://discussion.fedoraproject.org/t/guide-to-interacting-with-this-site-b...
It's not perfect, but it's better than most other forum software's email interfaces.
Glad to see it is a way of interacting with the forum.
I enjoyed Fedora being on mailing lists, nothing ever came close to the threading of emails. I was not getting lost in threads of conversation while still being under the umbrella topic, no need to open who knows how many links to read all tangents.
I appreciate your perspective, feedback, and willingness to try this out!
I will always remember email communications, but the only constant is change, and I am a very much "let's try and see" kind of person.
If this is going to end up as a positive result for the Fedora Project as a whole, well, I'll be happy about that.
Thanks, Jarek
On Fri, Apr 21, 2023 at 07:39:17PM +0200, Jaroslav Prokop wrote:
I am, luckily, not paid to read forums with no threading. IMO, a stream of posts with mentions of previous posts is not threading. Threading begins and ends on new topic posts AFAICT on discourse.
It's not presented as a tree, but there _are_ threads of replies.
Heh, sounds like a fun side project to try to transform it into a tree structure.
If you want to make Jonathan Corbet happy, make that tree structure be then served via NNTP. :)
Example:https://discussion.fedoraproject.org/t/future-of-encryption-in-fedora-deskto...
Finally, noticed what it does, it made me a bit confused as the first response was the same as in the "global" flow of the topic, but the message under it changed. I think that it should be better visible that they are actually replies.
It seems to hide other replies and only show those that are part of the "thread". Do they accept RFEs? :)
They do -- post at https://meta.discourse.org/.
I think enhancing the visibility after I expand replies for the posts in the "thread" would be better.
This particular thing could possibly be done via a theme component (see https://meta.discourse.org/t/about-the-theme-component-category/232731), which is a kind of lightweight plugin for the client-side. (These are very easy to install into a given site, unlike weightier server-side plugins.)
Or maybe even just some CSS. What exactly did you have in mind?
On 4/21/23 20:44, Matthew Miller wrote:
On Fri, Apr 21, 2023 at 07:39:17PM +0200, Jaroslav Prokop wrote:
I am, luckily, not paid to read forums with no threading. IMO, a stream of posts with mentions of previous posts is not threading. Threading begins and ends on new topic posts AFAICT on discourse.
It's not presented as a tree, but there _are_ threads of replies.
Heh, sounds like a fun side project to try to transform it into a tree structure.
If you want to make Jonathan Corbet happy, make that tree structure be then served via NNTP. :)
Sounds like a challenge! I'll put it on my personal project backlog ;)
Example:https://discussion.fedoraproject.org/t/future-of-encryption-in-fedora-deskto...
Finally, noticed what it does, it made me a bit confused as the first response was the same as in the "global" flow of the topic, but the message under it changed. I think that it should be better visible that they are actually replies.
It seems to hide other replies and only show those that are part of the "thread". Do they accept RFEs? :)
They do -- post athttps://meta.discourse.org/.
I think enhancing the visibility after I expand replies for the posts in the "thread" would be better.
This particular thing could possibly be done via a theme component (see https://meta.discourse.org/t/about-the-theme-component-category/232731), which is a kind of lightweight plugin for the client-side. (These are very easy to install into a given site, unlike weightier server-side plugins.)
Or maybe even just some CSS. What exactly did you have in mind?
I think it is a bug in UI/UX. I click "replies", I see *something* changed in the layout, but I have trouble identifying what exactly. Visually, if there are more than 2 or even 1 reply, then I just see with peripheral vision something had changed, but when I look closer I have trouble seeing a difference, IOW, I have trouble identifying what I did when I clicked the button.
Either moving the replies a bit to a side or maybe adding something more visual, like a blue stripe on the threaded replies would be helpful IMO.
I can play around with the plugin, see what feels how, and come back to Discourse with suggestions, thanks for the resources!
(Firstly I'll have to confirm with a more "vanilla" install, to see if the problem might be my default dark mode environment Discourse switched into, that's a problem for monday me though :) )
Regards, Jarek
On 4/21/23 19:39, Jaroslav Prokop wrote:
On 4/21/23 17:42, Matthew Miller wrote:
On Fri, Apr 21, 2023 at 10:50:42AM +0200, Jarek Prokop wrote:
also drives us towards more scattered communications. Our infamous mega-threads are not really effective for getting to community consensus, and tend to bring out the worst in us.
Passionate people generate passionate discussion.
The only thing you will gain by a forum is that at the point the message will not be deemed appropriate, it will probably be deleted or "beatified" by the mod team. The passion from our human nature will not go away with a platform change.
That's true -- and I'm not looking to get rid of passion, or silence opinions. But when something is _really_ out of line (often written in the heat of the moment), it's better to have options to ... as you say, beautify* the conversation. That makes it better for other people participating, and better for the person who has a chance to make their point in a more constructive way.
- also, to fix typos :)
Oh, probably an important related feature I noticed after looking at Chris Adams' response, I had a small concern about people changing messages too radically, where the conversation will then lose meaning, the software actually supports history and colorful diffs.
[snip]
A discussion to a technical change, for me, will forever be in a ticket. No matter the "wider discussion platform" projects will always have bug trackers where one can create a ticket.
Of course. That's not what I'm talking about. Consider for example this: https://pagure.io/fesco/issue/2817. That's not about the technical decision itself -- it's an branch of the conversation that should have been here.
biased towards those for whom it is working just fine. But, core Fedora development discussion can’t be limited to that ever-shrinking group. Consider who isn’t here. The problems are real, and the trend isn’t in a good direction.
But, is it shrinking due to a platform, or something other?
I don't think Fedora contribution and activity overall are shrinking. And I'm quite convinced that the platform is part of it.
It makes me want to try discourse out, not saying I'll stick around,
I'm glad to hear that.
I am, luckily, not paid to read forums with no threading. IMO, a stream of posts with mentions of previous posts is not threading. Threading begins and ends on new topic posts AFAICT on discourse.
It's not presented as a tree, but there _are_ threads of replies.
Heh, sounds like a fun side project to try to transform it into a tree structure.
If you see something like "2 replies" under a particular post, you can click that and the view will be restricted to just those replies, which you can then follow further.
Example:https://discussion.fedoraproject.org/t/future-of-encryption-in-fedora-deskto...
Finally, noticed what it does, it made me a bit confused as the first response was the same as in the "global" flow of the topic, but the message under it changed. I think that it should be better visible that they are actually replies.
It seems to hide other replies and only show those that are part of the "thread". Do they accept RFEs? :) I think enhancing the visibility after I expand replies for the posts in the "thread" would be better.
But also, yes — when something really diverges in Discourse, it should be a new topic. A moderator can move things after the fact (like I did with https://discussion.fedoraproject.org/t/getting-systemd-homed-working-properl...) but even better, when replying, you can create a linked topic. See https://discussion.fedoraproject.org/t/site-tip-create-linked-topics-for-dee...
But I'd be happier if there was some tangible metric how to measure if we got more *related to the topic* engagement. I would hate to see 20 "+1" posts from "random" users counted towards "it is better now".
That's reasonable. Do you have suggestions for a good metric?
I'm afraid none that could be automated, but I am not one strong on metrics. I'll just throw out some ideas:
- Number of unique contributors
- How many unique posts these contributors interacted with
- "quality" of the post. I think one could go by the length and
verbosity of the post. E.g. "Yeah seems like a good change" is not as valuable as a deeper dive/analysis into a hypothetical problematic. (especially if we consider that the platform has +1 equivalents in reactions :)) 4. Number/frequency of interactions.
Maybe a combination of 1. and 4. would have value.
But we can worry about that a bit later than "right now".
In Project Discussion, each different Fedora team can have its own tag, and you can subscribe to those that you’re interested in. Cross-posting is easy: tag a post with multiple teams.
I'd be interested in having a kind of "crossroad sign", to direct me towards tags what I would care about from a packager perspective. Not happy about this change, but it would make my experience a bit better...
There's a big _index_ athttps://discussion.fedoraproject.org/tags, but that's probably bit much (while at the same time not containing enough description). What would this "sign" ideally look like to you?
Usually we put these into "might want to follow" in various resources.
Perhaps a relevant tag list in the Fedora documentation? There are various onboarding and "must read" level stuff where I think we point to Fedora devel mailing list (or equivalent for SIGs).
Basically a recommendation of "follow new proposals with tag #proposals, packaging questions with #packaging ...... tags on Fedora Discussion"
TBH, I enjoy simplicty, in time of writing, I was imagining something akin to a wiki page with a few lines on recommended tags.
I agree that we need some kind of recommended presets.
Matthew has a nice overview of the forum structure at
https://discussion.fedoraproject.org/t/navigating-fedora-discussion-tags-cat...
But no one wants to read guides to just be able to communicate. Neither newbies, nor the old-timers :)
So we need something very simple as well, like: if you are used to fedora-devel, here is the recommended setup (theme?, subscriptions, set of defaults) for you. You can get more from the Discourse, but you don't have to.
Same for other active mailing lists.
Also, personally I am waiting for the Discourse sidebar. You can see it on the main forum https://meta.discourse.org/ but it is not finished and therefore not rolled out to Fedora instance yet.
While I like that one thread can be tagged via different tags, I still prefer to think about forum as a collection of rooms, not a single flow of all things to which I am subscribed to. And I can choose to enter the dedicated room or not to enter.
I can right now click on a specific tag to get the list of things relevant to it, but it is not the same. Sidebar provides that always visible room list, which we are used to on IRC, Matrix, Slack or in the mail client, and therefore makes this "room structure" much more obvious. And I think it would help a lot to improve the navigation in the current unstructured "tag cloud".
...
And now I think that maybe we should setup a set of tags with -room suffux, as it is describes it better than the -team which we seem to be using now.
#devel-room, #qa-room, #packaging-room, etc.
That said, it is web-first software. (Or mobile, if that’s your thing.) That requires some adjustment, I know. I hope opening up a Fedora Discussion tab – or keeping one open — becomes an easy habit.
If I was a volunteer that's the thing I'd remember once in a blue moon that it even exists. But I guess that's just person to person :).
There _are_ email notifications, and you can interact by replying to them. (You can even +1 or <3.)
There is also a "digest" mail sent automatically if you're not active, showing active topics possibly of interest, which can serve as a more-frequent-than-blue-moon reminder. (You can turn this off, of course.)
That is good to know.
As a person in my early 20s, I hate how everything is becoming web centric and no one can convince me to feel otherwise. While I am hearing from varying people around me, how it must be bad using email, it provides client-side filtering unparalleled by any platform that I used in the past.
It's fine, but it's no NNTP. That was really the best. :)
Do take a look at
https://discussion.fedoraproject.org/t/guide-to-interacting-with-this-site-b...
It's not perfect, but it's better than most other forum software's email interfaces.
Glad to see it is a way of interacting with the forum.
I enjoyed Fedora being on mailing lists, nothing ever came close to the threading of emails. I was not getting lost in threads of conversation while still being under the umbrella topic, no need to open who knows how many links to read all tangents.
I appreciate your perspective, feedback, and willingness to try this out!
I will always remember email communications, but the only constant is change, and I am a very much "let's try and see" kind of person.
If this is going to end up as a positive result for the Fedora Project as a whole, well, I'll be happy about that.
Thanks, Jarek
Matthew Miller wrote:
There _are_ email notifications,
You keep talking about "notifications". I don't want a notification saying "somebody said something; run our Javascript program to find out what it was". I want the actual message delivered to my mailbox.
But some people in this thread talk as if it's possible to get the actual messages by email. I even see some hints that a proper tree view might be possible. But many others say the email features of Discourse are no good. The overall picture is unclear and far from convincing.
and you can interact by replying to them.
You let the quarrel go this far before you even bothered to mention that? You're clearly not serious about selling the email features of Discourse.
You're trying to convince email users that your preferred communication program is better than their preferred communication program. Users of various email programs are saying no, the program I have chosen works better for my needs. Do you also waste time trying to convince Emacs users to switch to Vi?
Instead of trying to make everybody use the same Javascript program, what you should do is show how your preferred program implements the relevant standards to be interoperable with my preferred program, so that we can communicate while each using our respective programs. If it's not interoperable, then that's where the problem is.
So how complete is Discourse's email functionality actually? Can it be used as a mailing list server, or not?
Do take a look at
https://discussion.fedoraproject.org/t/guide-to-interacting-with-this-site-b...
Okay okay fine! I'll start up a browser in which the Javascript program even works and go read in the Javascript program about how I might not have to use the Javascript program.
So it says a "tag" is supposed to be sort of like a mailing list, but there's only a single global email address for starting new threads. There's no way to send to a specific tag. Thus it's impossible to post to a specific mailing-list-equivalent. How am I supposed to ensure that my messages reach the appropriate audience?
Maybe by replying? It's rather unclear how replies by email are handled, but I can guess that they're given the same tag as the message they reply to. If that's the case, then it almost seems like they *want* people to reply to an irrelevant thread instead of posting a new thread. I suppose a more likely explanation is that the email notifications are meant to draw users back to the Javascript program.
Either way, according to that post the answer is no, Discourse is not usable as a list server. Those who want to replace Mailman with Discourse should work on improving its email capabilities until it can be used as a list server.
Björn Persson
On Sun, Apr 23 2023 at 05:01:53 PM +0200, Björn Persson Bjorn@xn--rombobjrn-67a.se wrote:
But some people in this thread talk as if it's possible to get the actual messages by email. I even see some hints that a proper tree view might be possible. But many others say the email features of Discourse are no good. The overall picture is unclear and far from convincing.
You have to check one checkbox in settings if you want to get messages by mail. The mailing list mode is not amazing, but it works.
Il 23/04/23 17:01, Björn Persson ha scritto:
You let the quarrel go this far before you even bothered to mention that? You're clearly not serious about selling the email features of Discourse.
You're trying to convince email users that your preferred communication program is better than their preferred communication program. Users of various email programs are saying no, the program I have chosen works better for my needs. Do you also waste time trying to convince Emacs users to switch to Vi?
Instead of trying to make everybody use the same Javascript program, what you should do is show how your preferred program implements the relevant standards to be interoperable with my preferred program, so that we can communicate while each using our respective programs. If it's not interoperable, then that's where the problem is.
I'm not enthusiast about this proposed change too, but let's not go personal.
As I understand, it's not about Matthew personal preferences, it's about maintaining mailing list running needs manpower which Fedora is running out of. Switching to Discourse can probably relief some of the burden from fedora-infra guys.
Mattia
On Sun, Apr 23, 2023 at 04:54:48PM +0000, Mattia Verga via devel wrote:
Il 23/04/23 17:01, Björn Persson ha scritto:
You let the quarrel go this far before you even bothered to mention that? You're clearly not serious about selling the email features of Discourse.
You're trying to convince email users that your preferred communication program is better than their preferred communication program. Users of various email programs are saying no, the program I have chosen works better for my needs. Do you also waste time trying to convince Emacs users to switch to Vi?
Instead of trying to make everybody use the same Javascript program, what you should do is show how your preferred program implements the relevant standards to be interoperable with my preferred program, so that we can communicate while each using our respective programs. If it's not interoperable, then that's where the problem is.
I'm not enthusiast about this proposed change too, but let's not go personal.
100% agreed.
As I understand, it's not about Matthew personal preferences, it's about maintaining mailing list running needs manpower which Fedora is running out of. Switching to Discourse can probably relief some of the burden from fedora-infra guys.
I can't speak (or type) for Matthew, but I don't think that actually has much to do with this. I'd love to be running a newer/better supported mailman3 (and some folks have been hard at work to make that happen), but even if right now today we had a super clean, well supported mailman3 install for our lists, all the points in Matthews email would still apply. He didn't mention mailman3 in his post at all that I can see. I think we will be running mailing lists in some form for a long while still, just fewer people will be interacting with them.
kevin
On Sun, Apr 23, 2023 at 05:01:53PM +0200, Björn Persson wrote:
Matthew Miller wrote:
There _are_ email notifications,
You keep talking about "notifications". I don't want a notification saying "somebody said something; run our Javascript program to find out what it was". I want the actual message delivered to my mailbox.
But some people in this thread talk as if it's possible to get the actual messages by email. I even see some hints that a proper tree view might be possible. But many others say the email features of Discourse are no good. The overall picture is unclear and far from convincing.
Yes, you absolutely can interact with discourse via email (mostly). You will need to go to the website to setup things initially, but after that you should not need to.
and you can interact by replying to them.
You let the quarrel go this far before you even bothered to mention that? You're clearly not serious about selling the email features of Discourse.
You're trying to convince email users that your preferred communication program is better than their preferred communication program. Users of various email programs are saying no, the program I have chosen works better for my needs. Do you also waste time trying to convince Emacs users to switch to Vi?
Instead of trying to make everybody use the same Javascript program, what you should do is show how your preferred program implements the relevant standards to be interoperable with my preferred program, so that we can communicate while each using our respective programs. If it's not interoperable, then that's where the problem is.
So how complete is Discourse's email functionality actually? Can it be used as a mailing list server, or not?
It can send you emails for the things you tell it you want emails and you can sort and read and reply to them as you like.
Do take a look at
https://discussion.fedoraproject.org/t/guide-to-interacting-with-this-site-b...
Okay okay fine! I'll start up a browser in which the Javascript program even works and go read in the Javascript program about how I might not have to use the Javascript program.
So it says a "tag" is supposed to be sort of like a mailing list, but there's only a single global email address for starting new threads.
Yes, as noted, this is currently the setup for new threads/discussions due to spam. Possibly some better way can be implemented.
There's no way to send to a specific tag. Thus it's impossible to post to a specific mailing-list-equivalent. How am I supposed to ensure that my messages reach the appropriate audience?
Thats the weakest part of the setup for discourse by email right now I think. You can easily reply to existing posts just fine. It's just new threads/discussions that you need to either post from the web interface or send into the one address and have some moderator tag it and approve it.
Maybe by replying? It's rather unclear how replies by email are handled, but I can guess that they're given the same tag as the message they reply to. If that's the case, then it almost seems like they *want* people to reply to an irrelevant thread instead of posting a new thread. I suppose a more likely explanation is that the email notifications are meant to draw users back to the Javascript program.
You can reply to any email you get. It appears to use a reply-to thats a hash so it knows who you are and what post/discussion you are replying to so it can process your reply.
Either way, according to that post the answer is no, Discourse is not usable as a list server. Those who want to replace Mailman with Discourse should work on improving its email capabilities until it can be used as a list server.
Things can always be improved by "used as a list server" is not a spec of concrete proposal for functionality.
Anyhow, I invite you to give it a try. I am using the email gateway here and for the most part it works fine.
There's a bug I did hit where some of the categories seem to put \n in List-Id: headers, which is anoying to manage, but aside from that, all the emails come in, I can reply. We could probibly come up with some better way to start new topics/discussions, but in the mean time either the web interface or just sending an email and adding a 'please tag this as #foo' should work I would think.
kevin
Kevin Fenzi wrote:
We could probibly come up with some better way to start new topics/discussions
Yes I think I can come up with a better way. Give each tag its own email address, like a mailing list. That was very easy to come up with.
You can write the tag after the plus sign if that makes it easier to implement. Instead of "fedoraproject+newtopic@discoursemail.com" the address could be "fedoraproject+devel@discoursemail.com" or maybe "fedoraproject+devel/newtopic@discoursemail.com". Or some other format. The local-part can be structured any way the Discourse developers like, as long as it's at most 64 bytes and adheres to the dot-atom-text syntax in RFC 5322.
Björn Persson
On Sun, Apr 23, 2023 at 11:21:58PM +0200, Björn Persson wrote:
Kevin Fenzi wrote:
We could probibly come up with some better way to start new topics/discussions
Yes I think I can come up with a better way. Give each tag its own email address, like a mailing list. That was very easy to come up with.
I think you mean each category? But you may want multiple tags on a post... ie,
https://discussion.fedoraproject.org/t/fedora-strategy-2028-focus-area-revie... is posted in the 'Project Discussion' category with #council and #strategy2028 tags.
You can write the tag after the plus sign if that makes it easier to implement. Instead of "fedoraproject+newtopic@discoursemail.com" the address could be "fedoraproject+devel@discoursemail.com" or maybe "fedoraproject+devel/newtopic@discoursemail.com". Or some other format. The local-part can be structured any way the Discourse developers like, as long as it's at most 64 bytes and adheres to the dot-atom-text syntax in RFC 5322.
But that also doesn't solve the spam problem... anyone could send to those addresses, and indeed spammers will. ;(
But perhaps this could be useful with some other way to autenticate posts. Or perhaps there could be some kind of spam detection and require 'spammy' looking posts to be moderated?
I'm not a discourse maintainer tho, so I have no idea how open they would be to implementing this.
kevin
Dne 24. 04. 23 v 2:32 Kevin Fenzi napsal(a):
On Sun, Apr 23, 2023 at 11:21:58PM +0200, Björn Persson wrote:
Kevin Fenzi wrote:
We could probibly come up with some better way to start new topics/discussions
Yes I think I can come up with a better way. Give each tag its own email address, like a mailing list. That was very easy to come up with.
I think you mean each category? But you may want multiple tags on a post...
You could send it to multiple email addresses, right?
Vít
In my opinion, in this kind of big change, it's important for us to make things reversible, and to consider a possibility for us to go back.
I think it's better for us to keep the infra of the mailing list for a while. When we migrate the devel@ to the discourse, we should still keep to manage other mailing lists without migrating to the discourse. It's risky to migrate every mailing list at the same time or in a short period.
On 4/23/23 18:32, Kevin Fenzi wrote:
You can write the tag after the plus sign if that makes it easier to implement. Instead of"fedoraproject+newtopic@discoursemail.com" the address could be"fedoraproject+devel@discoursemail.com" or maybe "fedoraproject+devel/newtopic@discoursemail.com". Or some other format. The local-part can be structured any way the Discourse developers like, as long as it's at most 64 bytes and adheres to the dot-atom-text syntax in RFC 5322.
But that also doesn't solve the spam problem... anyone could send to those addresses, and indeed spammers will. ;(
I have not used the forum mailing lists so I don't know if they already include the following: Perhaps in each email include the url corresponding to the forum post, so that if you want to answer it, you have to do it from the forum, so the mailing lists can be read-only Gabrielo
Kevin Fenzi wrote:
On Sun, Apr 23, 2023 at 11:21:58PM +0200, Björn Persson wrote:
Kevin Fenzi wrote:
We could probibly come up with some better way to start new topics/discussions
Yes I think I can come up with a better way. Give each tag its own email address, like a mailing list. That was very easy to come up with.
I think you mean each category?
I don't know Discourse but we're told that something called a tag is roughly equivalent to a mailing list. I suppose categories could have addresses too.
But you may want multiple tags on a post...
Like Vít said, you can send to multiple addresses. That's how you cross-post to multiple mailing lists. The Discourse server would then read all the addresses and apply all of those tags and/or categories to the post.
When there are multiple recipient addresses in the same domain, a well-behaved SMTP client is supposed to transmit a single copy of the message in a single SMTP session with multiple RCPT commands. Thus the Discourse server will receive only one copy.
It is however possible that some badly written program might mishandle such a message and send a separate copy to each recipient address. Each copy would then still contain the whole list of addresses in the To and CC fields. If the Discourse server would read the header fields and not just the SMTP envelope, then the copies would appear as duplicate posts, each with the full set of tags, not as separate posts with one tag each.
If duplicates would turn out to be a great nuisance, then the Discourse developers might want to add a deduplication feature. The Message-ID field would be useful for discovering duplicates, but deduplication should not be done based on the message ID alone. The full contents should be compared to ensure that the messages really are identical, in case some defective or malicious email client produces non-unique message IDs.
As you can see, it doesn't take any great inventions to do this. The email standards already contain the necessary features. They just need to be implemented, if the Discourse developers are serious about supporting interaction by email.
But that also doesn't solve the spam problem... anyone could send to those addresses, and indeed spammers will. ;(
We're told that only sender addresses associated with a Fedora account are allowed to send to the single global new-topic address. Obviously that would apply to the tag (and category) addresses too. That's analogous to reducing spam to mailing lists by accepting posts only from subscribers.
In what scenario do tag-specific new-topic addresses result in a worse spam problem than a single global new-topic address?
But perhaps this could be useful with some other way to autenticate posts.
I haven't seen spammers impersonate subscribers in the mailing lists. The occasional spam that gets into the mailing lists seems to be done by subscribing a disposable address and sending from that address.
If spammers would start putting in a legitimate user's address as sender to get the spam into mailing lists or Discourse, then there's DKIM. I have found DKIM by itself ineffective, as most of the spam is DKIM- signed now, but DKIM combined with a requirement for a known sender address should be sufficient authentication to stop spam. The spammer would at least have to actually send from the same domain as the user they impersonate.
For registered users whose email provider doesn't sign their messages with DKIM, a verification message could be sent that they have to reply to, like when signing up for a mailing list but repeated for every post that isn't a reply. There's also OpenPGP/MIME. But I rather doubt that such measures will be needed just to fight spam. Strong authentication is for preventing more targeted attacks than spam.
Björn Persson
On Tue, Apr 25, 2023 at 12:12:05AM +0200, Björn Persson wrote:
Kevin Fenzi wrote:
On Sun, Apr 23, 2023 at 11:21:58PM +0200, Björn Persson wrote:
Kevin Fenzi wrote:
We could probibly come up with some better way to start new topics/discussions
Yes I think I can come up with a better way. Give each tag its own email address, like a mailing list. That was very easy to come up with.
I think you mean each category?
I don't know Discourse but we're told that something called a tag is roughly equivalent to a mailing list. I suppose categories could have addresses too.
I'm not sure I would say that... I guess there's no 100% equivalents here.
categories are like "Project Discussion" or "Ask Fedora" and tags can be any number on any thread.
ie, under "Project Discussion" there's a post about the new website fronpage revamp: https://discussion.fedoraproject.org/t/fedora-workstation-front-page-revamp-... that has tags "mindshare websites-and-apps-team design-team marketing-team"
You can watch a category, or a tag or multiple tags.
I guess it depends on the level of things you want to get.
But you may want multiple tags on a post...
Like Vít said, you can send to multiple addresses. That's how you cross-post to multiple mailing lists. The Discourse server would then read all the addresses and apply all of those tags and/or categories to the post.
When there are multiple recipient addresses in the same domain, a well-behaved SMTP client is supposed to transmit a single copy of the message in a single SMTP session with multiple RCPT commands. Thus the Discourse server will receive only one copy.
It is however possible that some badly written program might mishandle such a message and send a separate copy to each recipient address. Each copy would then still contain the whole list of addresses in the To and CC fields. If the Discourse server would read the header fields and not just the SMTP envelope, then the copies would appear as duplicate posts, each with the full set of tags, not as separate posts with one tag each.
If duplicates would turn out to be a great nuisance, then the Discourse developers might want to add a deduplication feature. The Message-ID field would be useful for discovering duplicates, but deduplication should not be done based on the message ID alone. The full contents should be compared to ensure that the messages really are identical, in case some defective or malicious email client produces non-unique message IDs.
Sure, thats all possible.
As you can see, it doesn't take any great inventions to do this. The email standards already contain the necessary features. They just need to be implemented, if the Discourse developers are serious about supporting interaction by email.
well, as you well know, coming up with ideas on how things could work is often the easy part. :) I have no idea how willing they would be to work on this... but you can ask on https://meta.discourse.org/
But that also doesn't solve the spam problem... anyone could send to those addresses, and indeed spammers will. ;(
We're told that only sender addresses associated with a Fedora account are allowed to send to the single global new-topic address. Obviously
I don't think thats the case at all. Currently I think anyone can send, it just gets moderated. But I would defer to Matthew here...
that would apply to the tag (and category) addresses too. That's analogous to reducing spam to mailing lists by accepting posts only from subscribers.
It's worth noting that if you get emails from discourse the reply-to is set to a hash so it knows who you are and what you are replying to so it can insert it in.
In what scenario do tag-specific new-topic addresses result in a worse spam problem than a single global new-topic address?
Currently as far as I know if you send in, you need to either be using a reply-to that has the right hash or sending to the global email which will be moderated. If we unmoderated the global address it would be the same spam problem as new-topic ones would have (although that would help solve the topic problem).
But perhaps this could be useful with some other way to autenticate posts.
I haven't seen spammers impersonate subscribers in the mailing lists. The occasional spam that gets into the mailing lists seems to be done by subscribing a disposable address and sending from that address.
Usually yes. I have seen impersonations in the past. It doesn't seem to be as common anymore.
If spammers would start putting in a legitimate user's address as sender to get the spam into mailing lists or Discourse, then there's DKIM. I have found DKIM by itself ineffective, as most of the spam is DKIM- signed now, but DKIM combined with a requirement for a known sender address should be sufficient authentication to stop spam. The spammer would at least have to actually send from the same domain as the user they impersonate.
Perhaps. I don't know if discourse can implement some kind of incoming checks on emails. Matthew?
For registered users whose email provider doesn't sign their messages with DKIM, a verification message could be sent that they have to reply to, like when signing up for a mailing list but repeated for every post that isn't a reply. There's also OpenPGP/MIME. But I rather doubt that such measures will be needed just to fight spam. Strong authentication is for preventing more targeted attacks than spam.
Yeah, thats another possible solution... just require a ack/confirm to post. That would stop a lot (but not all) spammers.
Also throttling could be possible. Only X new posts from a address in Y time.
Anyhow, we should probibly try and move this upstream and see if they are willing to work on any of this, or have other plans of their own. ;)
Thanks for the constructive discussion!
kevin
On 25-04-2023 18:58, Kevin Fenzi wrote:
As you can see, it doesn't take any great inventions to do this. The email standards already contain the necessary features. They just need to be implemented, if the Discourse developers are serious about supporting interaction by email.
well, as you well know, coming up with ideas on how things could work is often the easy part. 😄 I have no idea how willing they would be to work on this... but you can ask onhttps://meta.discourse.org/
I put in a request for getting the plain text e-mail formatting of quotes fixed:
https://meta.discourse.org/t/format-quotes-in-plain-text-e-mail-correctly/26...
Feel free to chime in and/or add your own requests.
-- Sandro
* Kevin Fenzi:
Yes, you absolutely can interact with discourse via email (mostly).
This option has been disabled in the Fedora instance, though. There should be “Mailing list mode“ and “Enable mailing list mode” at the end of this page:
https://discussion.fedoraproject.org/u/fweimer/preferences/emails
But in my case at least, it's missing.
Thanks, Florian
On Tue, Apr 25 2023 at 11:32:14 AM +0200, Florian Weimer fweimer@redhat.com wrote:
This option has been disabled in the Fedora instance, though. There should be “Mailing list mode“ and “Enable mailing list mode” at the end of this page:
https://discussion.fedoraproject.org/u/fweimer/preferences/emails
But in my case at least, it's missing.
Seconding this question. What's up with this?
The mailing list mode is not amazing, but it is adequate. Without it, participation in Fedora devel discussions will plummet. At least there's no chance I'll manually browse the web forum to see what's going on.
Michael
On Tue, Apr 25, 2023 at 10:24:37AM -0500, Michael Catanzaro wrote:
On Tue, Apr 25 2023 at 11:32:14 AM +0200, Florian Weimer fweimer@redhat.com wrote:
This option has been disabled in the Fedora instance, though. There should be “Mailing list mode“ and “Enable mailing list mode” at the end of this page:
https://discussion.fedoraproject.org/u/fweimer/preferences/emails
But in my case at least, it's missing.
Seconding this question. What's up with this?
The mailing list mode is not amazing, but it is adequate. Without it, participation in Fedora devel discussions will plummet. At least there's no chance I'll manually browse the web forum to see what's going on.
As Matthew said in an eariler post in this very thread: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org/...
"Mailing list mode" was a specific thing in earlier versions of Discourse — it sent a notification for every message posted. This is kind of like going to Hyperkitty and saying "subscribe me to all 600 lists". I don't recommend that. Instead, choose specific tags that you want to subscribe to, just as you would subscribe to individual mailing lists.
kevin
* Kevin Fenzi:
On Tue, Apr 25, 2023 at 10:24:37AM -0500, Michael Catanzaro wrote:
On Tue, Apr 25 2023 at 11:32:14 AM +0200, Florian Weimer fweimer@redhat.com wrote:
This option has been disabled in the Fedora instance, though. There should be “Mailing list mode“ and “Enable mailing list mode” at the end of this page:
https://discussion.fedoraproject.org/u/fweimer/preferences/emails
But in my case at least, it's missing.
Seconding this question. What's up with this?
The mailing list mode is not amazing, but it is adequate. Without it, participation in Fedora devel discussions will plummet. At least there's no chance I'll manually browse the web forum to see what's going on.
As Matthew said in an eariler post in this very thread: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org/...
Are you referring to the “in earlier versions of Discourse“, so it's just gone? With the expectation that we won't need it because we receive notifications for everything in the categories we “watch”?
Thanks, Florian
On Tue, Apr 25, 2023 at 07:25:06PM +0200, Florian Weimer wrote:
- Kevin Fenzi:
On Tue, Apr 25, 2023 at 10:24:37AM -0500, Michael Catanzaro wrote:
On Tue, Apr 25 2023 at 11:32:14 AM +0200, Florian Weimer fweimer@redhat.com wrote:
This option has been disabled in the Fedora instance, though. There should be “Mailing list mode“ and “Enable mailing list mode” at the end of this page:
https://discussion.fedoraproject.org/u/fweimer/preferences/emails
But in my case at least, it's missing.
Seconding this question. What's up with this?
The mailing list mode is not amazing, but it is adequate. Without it, participation in Fedora devel discussions will plummet. At least there's no chance I'll manually browse the web forum to see what's going on.
As Matthew said in an eariler post in this very thread: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org/...
Are you referring to the “in earlier versions of Discourse“, so it's just gone? With the expectation that we won't need it because we receive notifications for everything in the categories we “watch”?
Thats my understanding, yes. I could be wrong...
kevin
On Tue, 2023-04-25 at 10:42 -0700, Kevin Fenzi wrote:
On Tue, Apr 25, 2023 at 07:25:06PM +0200, Florian Weimer wrote:
- Kevin Fenzi:
On Tue, Apr 25, 2023 at 10:24:37AM -0500, Michael Catanzaro wrote:
On Tue, Apr 25 2023 at 11:32:14 AM +0200, Florian Weimer fweimer@redhat.com wrote:
This option has been disabled in the Fedora instance, though. There should be “Mailing list mode“ and “Enable mailing list mode” at the end of this page:
https://discussion.fedoraproject.org/u/fweimer/preferences/emails
But in my case at least, it's missing.
Seconding this question. What's up with this?
The mailing list mode is not amazing, but it is adequate. Without it, participation in Fedora devel discussions will plummet. At least there's no chance I'll manually browse the web forum to see what's going on.
As Matthew said in an eariler post in this very thread: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org/...
Are you referring to the “in earlier versions of Discourse“, so it's just gone? With the expectation that we won't need it because we receive notifications for everything in the categories we “watch”?
Thats my understanding, yes. I could be wrong...
I believe the upthread discussion said that in recent versions of Discourse it is disabled (server-side) by default, but there is a (again server-side) option to turn it on again.
How dare you - I'm glad you did :)
Even though I'm a "mail/mailing list guy" using TUI MUAs, I found myself turning delivery off on many high volume MLs where the volume does not correspond to my contributor's frequency. I even read fedora-devel via hyperkitty's web interface, which is really suboptimal. So I see both the value and the problem with MLs. A different transport like public-inbox may help me but not many others.
In any case, we have quite a fragmentation right now with the MLs, forum (discourse), IRC, Matrix, plus tickets on various platforms (bz, dist-git, pagure, gitlab) some of which offer teams and discussions, too. Choice is good, fragmentation is not because it makes it hard to know: - Where can I reach whom? - Where can I discuss what?
So I'm really all for reducing that fragmentation, and it can be made to work as a community decision only (community discussion that you started, whatever committee's decision). Ideally, we reduce the platforms to a few which still allow choices about how to participate (clickery vs tui, poll vs push/notify). More technically oriented folks will be more capable to adapt technically (than "pure users") but less willing to communicate by clicking around in a web browser. A platform analysis in this regard could support that.
On Fri, 2023-04-21 at 09:11 +0000, Michael J Gruber wrote:
In any case, we have quite a fragmentation right now with the MLs, forum (discourse), IRC, Matrix, plus tickets on various platforms (bz, dist-git, pagure, gitlab) some of which offer teams and discussions, too. Choice is good, fragmentation is not because it makes it hard to know:
- Where can I reach whom?
- Where can I discuss what?
This is true, but I think it's interestingly worth noting we've pretty much *always* had the same fragmentation.
For instance, as I recall, back when I joined in 2009, we had:
* Mailing lists * Forum (fedoraforum - this is/was not official, but commonly used, and illustrates a point that if people want some form of communication, they'll use it, even if we don't have an "official" one; see also Fedora discord and Fedora telegram) * IRC (the addition of Matrix hasn't fragmented things much really, as IRC and Matrix are bridged in both directions) * Bugzilla * Trac (playing the approximate current role of Pagure issues) * Various platforms for "upstreams", like Sourceforge (playing the current role of Pagure source tracking, gitlab, github etc.) * Usenet (alt.os.linux.fedora. This is *still* around, and some wonderful genius whose hand I want to shake has set up a bot that forwards Fedora Magazine posts to it. The last post by someone other than a bot was from "nw", on January 20. Until today. Now it's from me!)
I'm not sure it's possible to achieve zero- or low-fragmentation. Fragmentation just seems to...*happen*. If you build one system that does everything it's too big and unwieldy and people use more targeted systems for 'simplicity'.
On 4/20/23 5:20 PM, Matthew Miller wrote:
It’s time to transform the Fedora devel list into something new
As someone that read this list 99% ot the time and 1% for replying, I have no problem with having a web based forum for discussions, but that kind of software if always awful for my kind of usage.
I like the mailing list methos because I get all the threads and I see what I wish to read in order to be informed of the general direction of the project and ignore the ones I don't want to, with a simple responsive UI, that even works when I am offline.
I know that there are people that love web based forums (I don't), but in order to not leave people behind with a lot of experience in the project, for new ones, maybe instead of only proposing a change, maybe proposing a change with good email integration is better.
Discourse default email integration is bad IMHO. The model I wish to have with email integration are the tools that were installed for the OpenJDK - GitHub integration. in that mailing list, you never notice you are reading a bunch of emails that comes from GitHub comments and I can pretty much read the entire discussions from my email client, without ever using a web browser.
On 4/21/23 8:17 AM, Robert Marcano wrote:
On 4/20/23 5:20 PM, Matthew Miller wrote:
It’s time to transform the Fedora devel list into something new
As someone that read this list 99% ot the time and 1% for replying, I have no problem with having a web based forum for discussions, but that kind of software if always awful for my kind of usage.
I like the mailing list methos because I get all the threads and I see what I wish to read in order to be informed of the general direction of the project and ignore the ones I don't want to, with a simple responsive UI, that even works when I am offline.
I know that there are people that love web based forums (I don't), but in order to not leave people behind with a lot of experience in the project, for new ones, maybe instead of only proposing a change, maybe proposing a change with good email integration is better.
Discourse default email integration is bad IMHO. The model I wish to have with email integration are the tools that were installed for the OpenJDK - GitHub integration. in that mailing list, you never notice you are reading a bunch of emails that comes from GitHub comments and I can pretty much read the entire discussions from my email client, without ever using a web browser.
On 4/20/23 5:20 PM, Matthew Miller wrote:
That said, it is web-first software. (Or mobile, if that’s your thing.) That requires some adjustment, I know. I hope opening up a Fedora Discussion tab – or keeping one open — becomes an easy habit.
I forgot to add. Discourse is worse for mobile users, the "app" was just an embedded web browser, the only thing efficient it did was to get a notification and remembering your session, everything else is worse that writing an email on a email client for phones.
On Fri, Apr 21, 2023 at 08:27:06AM -0400, Robert Marcano via devel wrote:
I forgot to add. Discourse is worse for mobile users, the "app" was just an embedded web browser, the only thing efficient it did was to get a notification and remembering your session, everything else is worse that writing an email on a email client for phones.
For what it's worth, I use Discourse on my phone all the time. As you say, the app is just a thin wrapper around the mobile web UI, but I like that UI well enough.
And the wrapper is nice as a "hub" for various discourse sites, where I can see all of the notifications in one place.
On 4/20/23 17:20, Matthew Miller wrote:
Most of the conversation was under the subject “Schedule for Tuesday's FESCo Meeting (2023-01-03)”, because everything started as a reply to that. That’s pretty easy to overlook. It’s possible for replies to change the subject when replying, but that can’t be done retroactively, and then isn’t consistent (and it breaks threading in Gmail, too).
At the risk of causing at least one of the problems that Matthew pointed out, this is the most important one for me: lots of people say "I decide what to read based on what I see in my email client", but when the subject of the emails doesn't reflect their contents, that's a losing proposition. I completely ignored the thread in question until a *different* thread referred to it and I learned about the conversation going on there.
On another note: I've been an active Discourse user since the 'early adopter' days, and frequent at least six Discourse sites of varying content volumes. I've learned good ways to stay on top of what is going on without having to open each site on a daily basis, and while it's not as low-effort as mailing lists are, I've found the benefits to be worth the increased effort. In fact many of the replies in this thread have complaints/concerns about aspects of Discourse usage which have been 'solved' already, but the person making the complaint/voicing the concern just isn't aware of the solutions available yet... so for those people I encourage you to ask for assistance learning how to address the problems you perceive, instead of claiming the problems can't be solved.
A case in point is the statement "it's so hard to know what's new": I struggled with this as well, until I learned about configuring the behavior of the 'New' tab in Discourse, and aggressively filtering out categories and tags I do not care about. Now I can click 'refresh' while looking at the 'New' tab on a highly active site and I will see *all* of the New topics that are of potential interest to me. In the end this is very similar to taking the same actions in a local mail client, but more capable since it is based on metadata about the topics and not just the often-incorrect subject lines in email threads.
* Kevin P. Fleming [21/04/2023 10:03] :
lots of people say "I decidewhat to read based on what I see in my email client", but when the subject of the emails doesn't reflect their contents, that's a losing proposition.
I agree there's a huge lack of netiquette in Fedora's mailing lists, with wholesale quoting, top-posting, subjects not being updated, etc but changing mediums seems far more expensive than asking people to post emails that are easier to read.
Emmanuel
On Fri, Apr 21, 2023 at 04:55:00PM +0200, Emmanuel Seyman wrote:
I agree there's a huge lack of netiquette in Fedora's mailing lists, with wholesale quoting, top-posting, subjects not being updated, etc but changing mediums seems far more expensive than asking people to post emails that are easier to read.
Not to mention these problems won't go away just because it's now hosted on a web page..
- Solomon
On Fri, 2023-04-21 at 11:00 -0400, Solomon Peachy via devel wrote:
On Fri, Apr 21, 2023 at 04:55:00PM +0200, Emmanuel Seyman wrote:
I agree there's a huge lack of netiquette in Fedora's mailing lists, with wholesale quoting, top-posting, subjects not being updated, etc but changing mediums seems far more expensive than asking people to post emails that are easier to read.
Not to mention these problems won't go away just because it's now hosted on a web page..
Well, not because it's hosted on a web page, but a more structured system *does* address a lot of these problems. A lot of what you're referring to as "netiquette" is really about putting the onus on each individual user to do stuff that a more sophisticated system could handle for them, and a system like discourse *does* handle those. You can't top-post on discourse. Renaming topics works in a much saner way. You can deal with 'cross-posting' in a much saner and more flexible way. Quoting is always a problem everywhere, but that's only *one* thing, at least. :D
On Friday, 21 April 2023 at 20:17, Adam Williamson wrote:
On Fri, 2023-04-21 at 11:00 -0400, Solomon Peachy via devel wrote:
On Fri, Apr 21, 2023 at 04:55:00PM +0200, Emmanuel Seyman wrote:
I agree there's a huge lack of netiquette in Fedora's mailing lists, with wholesale quoting, top-posting, subjects not being updated, etc but changing mediums seems far more expensive than asking people to post emails that are easier to read.
Not to mention these problems won't go away just because it's now hosted on a web page..
Well, not because it's hosted on a web page, but a more structured system *does* address a lot of these problems. A lot of what you're referring to as "netiquette" is really about putting the onus on each individual user to do stuff that a more sophisticated system could handle for them, and a system like discourse *does* handle those. You can't top-post on discourse.
You don't have threaded replies, either. And quoting the post you're replying to is entirely non-intuitive. It took me a while to discover that you have to click on the "speech bubble" icon inside the editor that opens after you click reply.
Renaming topics works in a much saner way. You can deal with 'cross-posting' in a much saner and more flexible way. Quoting is always a problem everywhere, but that's only *one* thing, at least. :D
Quoting and threading are the fundamental features in text-based non-realtime conversations, so switching to a medium that offers inferior quoting and no threading is questionable at best.
As can be deduced from my UA header, I'm firmly in the e-mail camp and, similar to others in this thread, I find web-based forums unintuitive and difficult to use efficiently because of lack of efficient keyboard control. Yes, there are keyboard shortcuts in Discourse, but they don't allow you to browse through and open expand highlighted threads, jump to Nth thread/message, etc.
Efficient navigation requires combined keyboard/mouse usage, which disrupts the experience.
Regards, Dominik
* Dominik Mierzejewski:
You don't have threaded replies, either. And quoting the post you're replying to is entirely non-intuitive. It took me a while to discover that you have to click on the "speech bubble" icon inside the editor that opens after you click reply.
In some instances at least, you can select the text in the fragment you are replying to, and a quote button will appear. I at least found that quite discoverable.
Renaming topics works in a much saner way. You can deal with 'cross-posting' in a much saner and more flexible way. Quoting is always a problem everywhere, but that's only *one* thing, at least. :D
Quoting and threading are the fundamental features in text-based non-realtime conversations, so switching to a medium that offers inferior quoting and no threading is questionable at best.
I tend to agree, but most clients only offer rudimentary support to show the metadata that's there. Nowadays, a lot of people pick a random message in a thread and reply to that, disregarding what that particular message contributed to the conversation. I can't really blame them because in the interfaces they are used do, it does not make a difference for presentation purposes. I'm not sure if nested threading in email ever went mainstream, to be honest. Discourse and similar tools may merely reflect that lack of familiarity.
I couldn't consume the volume of mail I process without it because nested threading still works on most of the technical lists I frequent, but I suppose it's a dying art.
Thanks, Florian
On Tuesday, 25 April 2023 at 12:30, Florian Weimer wrote:
- Dominik Mierzejewski:
You don't have threaded replies, either. And quoting the post you're replying to is entirely non-intuitive. It took me a while to discover that you have to click on the "speech bubble" icon inside the editor that opens after you click reply.
In some instances at least, you can select the text in the fragment you are replying to, and a quote button will appear. I at least found that quite discoverable.
You have to use your mouse to do it, which is too disrupting for me.
Renaming topics works in a much saner way. You can deal with 'cross-posting' in a much saner and more flexible way. Quoting is always a problem everywhere, but that's only *one* thing, at least. :D
Quoting and threading are the fundamental features in text-based non-realtime conversations, so switching to a medium that offers inferior quoting and no threading is questionable at best.
I tend to agree, but most clients only offer rudimentary support to show the metadata that's there. Nowadays, a lot of people pick a random message in a thread and reply to that, disregarding what that particular message contributed to the conversation. I can't really blame them because in the interfaces they are used do, it does not make a difference for presentation purposes.
That seems to be the sad reality of many of the web-based forums I sometimes visit, so yes, my anecdotal evidence confirms your observations.
I'm not sure if nested threading in email ever went mainstream, to be honest.
It depends on the mailing lists you are subscribed to. The technical ones tend to make use of threading efficiently.
Discourse and similar tools may merely reflect that lack of familiarity.
No argument there.
I couldn't consume the volume of mail I process without it because nested threading still works on most of the technical lists I frequent, but I suppose it's a dying art.
Call me a dinosaur. ;) I have no intention of dying out any time soon, though. ;)
Regards, Dominik
On Tue, Apr 25, 2023 at 12:38:46PM +0200, Dominik 'Rathann' Mierzejewski wrote:
In some instances at least, you can select the text in the fragment you are replying to, and a quote button will appear. I at least found that quite discoverable.
You have to use your mouse to do it, which is too disrupting for me.
There are keybindings!
Use j/k to scroll up and down (ah, vi, your legacy will live forever) and q to start a reply to that post with the selected post marked as q quote, and then trim in the editor.
If you have your browser's arrow-key control feature turned on (F7 in Firefox), you can use this in combination — j/k to move between posts, and then arrow keys over to shift-arrow select the part you want, and then again q to quote (or if it is a post you have permission to edit — your own, or a wiki post, or you're a mod — e to edit).
On 4/26/23 11:25, Matthew Miller wrote:
Use j/k to scroll up and down (ah, vi, your legacy will live forever)
According to Wikipedia, it's actually the ADM-3A:
https://en.wikipedia.org/wiki/ADM-3A#Legacy
I recently learned this, and find it fascinating ☺
On Fri, Apr 21, 2023 at 04:55:00PM +0200, Emmanuel Seyman wrote:
- Kevin P. Fleming [21/04/2023 10:03] :
lots of people say "I decidewhat to read based on what I see in my email client", but when the subject of the emails doesn't reflect their contents, that's a losing proposition.
I agree there's a huge lack of netiquette in Fedora's mailing lists, with wholesale quoting, top-posting, subjects not being updated, etc but changing mediums seems far more expensive than asking people to post emails that are easier to read.
I generally find the term "netiquette" to be quite unpleasant, because it is historically frequently applied in user hostile manner, where people repeatedly rebuke (new) contributors/participants who don't follow the "rules", which has the effect of discouraging engagement with the project.
Contributor guidelines are fine, but it is important to accept, expect and tolerate that people are *NOT* going to following the guidelines, for a variety of reasons.
IMHO the application of "netiquette" rules in a particular scenario is usually a sign that the tools being used are inadequate for the job they're be applied to.
With regards, Daniel
I suppose it's ironically suitable that my first attempt to send the following failed because I wasn't actually subscribed to the list...
Matthew Miller mattdm@fedoraproject.org writes:
We’re missing people
I will just make the point that, when you make this switch, you will be missing people as well. Projects that switch to forum systems, to a great extend, go dark for people who aren't immersed in them all the time. Keeping up with fedora-devel takes very little of my time; keeping up with Discourse instances is a much slower affair.
I know I'm a grumpy old-timer, but I see this move to forum systems as a powerful force putting projects into silos and isolating them from each other.
Now get off my lawn, it's time to go watch Lawrence Welk...:)
jon
On Fri, Apr 21, 2023 at 08:45:31AM -0600, Jonathan Corbet wrote:
I will just make the point that, when you make this switch, you will be missing people as well. Projects that switch to forum systems, to a great extend, go dark for people who aren't immersed in them all the time. Keeping up with fedora-devel takes very little of my time; keeping up with Discourse instances is a much slower affair.
What could we do to make that easier for you?
Matthew Miller mattdm@fedoraproject.org writes:
On Fri, Apr 21, 2023 at 08:45:31AM -0600, Jonathan Corbet wrote:
I will just make the point that, when you make this switch, you will be missing people as well. Projects that switch to forum systems, to a great extent, go dark for people who aren't immersed in them all the time. Keeping up with fedora-devel takes very little of my time; keeping up with Discourse instances is a much slower affair.
What could we do to make that easier for you?
A decent NNTP feed (or set of feeds) would be wonderful :)
Otherwise, I don't know. Invent a decent federated system that keeps the good parts of email while addressing the many bad parts?
I guess I should give the Discourse email mode another try. It was painful last time, but it's been a little while and may have improved. But even if it's perfect, dealing with a lot of accounts on project-specific walled gardens is always going to increase the pain level. Perhaps I'm just not made for the world we have built...
Thanks,
jon
On 4/21/23 18:55, Jonathan Corbet wrote:
Matthew Miller mattdm@fedoraproject.org writes:
On Fri, Apr 21, 2023 at 08:45:31AM -0600, Jonathan Corbet wrote:
I will just make the point that, when you make this switch, you will be missing people as well. Projects that switch to forum systems, to a great extent, go dark for people who aren't immersed in them all the time. Keeping up with fedora-devel takes very little of my time; keeping up with Discourse instances is a much slower affair.
What could we do to make that easier for you?
A decent NNTP feed (or set of feeds) would be wonderful :)
If there will be a switch to a web based forum, I will be one of the subscribers that will be lost. Not sure if anybody cares, but I'm pointing it out since the proposal aims at having more participants. I've subscribed on 24/12/2005 and all the emails are in my IMAP server, and accessed with Thunderbird.
Honestly, the NNTP feed is exactly what I would have answered too.
I understand the world changes, and the attraction of avatars, and emojis powered by piles of javascript is unstoppable to some people. But from a marketing point of view it might not be the right direction: Fedora is an old-style technical thing, there is no point in chasing Ubuntu, everybody liking that kind of world will surely prefer the original.
Last but not least, mailing lists are one of the ways we can try keeping email service decently available to everybody, resisting the trend to let only a few companies be "senders", so either you write from gmail to gmail or you have to buy a "mail sending" service to have your mails accepted by receivers. It is bad enough already.
Regards.
TLDR; I'm in the "please, not a web forum" camp, but I also feel like this is effectively a foregone conclusion, unfortunately.
As a maintainer of a small number of packages, I follow devel to keep up with changes affecting the distribution and occasionally chime in or find areas where I can help in some way.
I have gained an immense amount of knowledge about Fedora this way as well as a strong appreciation for the folks doing the heavy lifting.
I've tried to use discourse a bit and find it to be like every other web forum -- heavy to load and frustrating to work with. My reasons match what a number of other folks here have already said.
In particular, I don't like moving from an email discussion method where each participant can choose their own client and tools to work as they want to a web interface which is one-size-fits-all (or none, as often everyone loses something).
I'd be among those who are simply less informed and engaged in the Fedora development community if things move to a web forum. If there are a lot more gains in doing so, I wish you well in the endeavor.
Thanks,
On 20. 04. 23 23:20, Matthew Miller wrote:
I propose that we transition devel list, and eventually most of our mailing lists, to Fedora Discussion (our Discourse-powered forum).
Python recently transitioned from a mailing list (python-dev) to Discourse (discuss.python.org).
Challenges I don't know how to deal with:
1. Python discussions are no longer at the same place as all the other discussions (my Thunderbird). This is a big one for me. I cannot primarily use the Discourse web interface on discuss.python.org because I simply won't go there. I suppose that if Fedora switches the devel mailing list to Discourse, I will have to go there, considering how often I post to the devel list. However, In suspect some more casual contributors will not go there adn we'll lost them or they will loose track od what's happening in Fedora.
A partial solution to this is to use the "mail list mode" and disable topics I don't care about. Which is possible, but brings me to...
2. When I read the content in the web app, it does not mark the related emails read. I can either read everything via email (which frankly has a worse readability than the web app) or go read it on web and than manually mark my emails as read. So far, this has been tedious for me so far and I repeatedly give up and mark the entire Python Discourse folder as read when the number reaches 10k unread emails. Even if I read everything via email, I cannot reply there so I still need to visit the thing in the web app to participate.
3. When I choose to use the mailing list approach, I can no longer distinguish "regular" email notifications (somebody replied to my comment or mentioned me) from the rest of the email traffic from the forum. This will apply to others as well; e.g. when I send an email to devel now, I can CC people I know need to see it -- on Discourse I can mention them, but they might not notice that if they receive an email of everything anyway.
So far, when Python switched from mailing list to Discourse I've noticed this:
Once I participate in a certain discussion, it somewhat works, as long as I remember to go check it out occasionally. But OTOH I miss out almost everything I don't actively participate in.
-----------
Perhaps I am a greybeard, but I am pretty much scared of this change in Fedora.
You say we miss people now. I say we will miss them then. Unfortunately, I don't know how to solve that. Maybe my fear is not justified, but it is real.
On Fri, Apr 21, 2023 at 08:53:26PM +0200, Miro Hrončok wrote:
- Python discussions are no longer at the same place as all the
other discussions (my Thunderbird). This is a big one for me. I cannot primarily use the Discourse web interface on discuss.python.org because I simply won't go there. I suppose that
I mentioned somewhere else around here that I use the Android "Discourse Hub" app as a way to keep up with this — I see notifications from Python, Home Asssitant, Flatpak, etc., even though I don't visit those sites daily.
I kind of miss this on the desktop. I wonder if a web app that does the same thing would help you too? As more and more projects use Discourse, that could become a sort of nexus for all of these gardens that Jonathan Corbet mentions.
- When I read the content in the web app, it does not mark the
related emails read. I can either read everything via email (which frankly has a worse readability than the web app) or go read it on web and than manually mark my emails as read. So far, this has been tedious for me so far and I repeatedly give up and mark the entire Python Discourse folder as read when the number reaches 10k unread emails. Even if I read everything via email, I cannot reply there so I still need to visit the thing in the web app to participate.
We have reply-by-email enabled. It is also possible to enable new topics by email, but that's vulernable to impersonation (and spam) so if we enable that there probably will be a moderation step.
- When I choose to use the mailing list approach, I can no longer
distinguish "regular" email notifications (somebody replied to my comment or mentioned me) from the rest of the email traffic from the forum. This will apply to others as well; e.g. when I send an email to devel now, I can CC people I know need to see it -- on Discourse I can mention them, but they might not notice that if they receive an email of everything anyway.
Can you use the Feedback-ID header to distinguish and highlight those personal notifications?
So far, when Python switched from mailing list to Discourse I've noticed this:
Once I participate in a certain discussion, it somewhat works, as long as I remember to go check it out occasionally. But OTOH I miss out almost everything I don't actively participate in.
Perhaps I am a greybeard, but I am pretty much scared of this change in Fedora.
You say we miss people now. I say we will miss them then. Unfortunately, I don't know how to solve that. Maybe my fear is not justified, but it is real.
Thank you. I appreciate the real-world feedback from the Python experience.
What about, instead of mailing list mode, enabling the Activity Summary email, and setting it to daily instead of weekly? That gives a reminder (and hopefully something interesting) but keeps the web site as primary for actual notifications and for keeping track of what's read and not read.
Matthew Miller wrote:
It is also possible to enable new topics by email, but that's vulernable to impersonation (and spam) so if we enable that there probably will be a moderation step.
Email signed with OpenPGP/MIME solves the impersonation and spam problems. A message could be allowed to bypass the moderation after it's verified that it's signed with the correct public key for an email address registered in a Fedora account.
DKIM also seems able to assert that a message is from a certain sender address, although almost all usage I've seen states only a domain name.
But if those new topics can't be sent to a mailing-list-equivalent, but just end up in some sort of "other" bucket, then it seems useless anyway.
Björn Persson
Not quoting anything in particular, just my opinion and a +1.
Being a solid Gen X'er I'm comfortable in both worlds. I will say that I've found the large volume of emails between Fedora and MythTV (though it's slowed in the last few years) occasionally overwhelming.
The first thing I do almost every morning (still in my PJs with a cup of coffee) is to go through all the threads, archive what I'm not interested in, but keep just in case, and read the others.
In my work world I frequently find myself in Discourse (or similar) forums, and I find it easy enough to find what I'm looking for, or if there's new threads or replies to threads I've posted to (usually at the top).
A few months ago I spent a few hours cleaning up old emails in gmail because google (the search engine) lies. I perfected a query to only return list emails older than 1 year but even though I checked "include all emails" it would only delete about 500 at a time. Took a while to work through my 10+ year history as a packager 500 emails at a time!
The reality is the future potential contributors are used to tools like Discourse. That's fine. I'll adapt.
Thanks, Richard
Hi Matthew,
On Thu Apr 20, 2023 at 17:20 -0400, Matthew Miller wrote
As it is, devel list is too much for many people to follow — people we’d like to have around.
I disagree. I find mailing lists much easier to deal with. They're all in one place, they're synced between my devices, and I can use whatever client I like. I can start writing a message on my phone, save it as a draft, and send it from my laptop. I can filter messages to my heart's content. Everything is threaded and I can choose to ignore the threads I'm uninterested in.
It's easy to skim the list and I can do so on my terms, as opposed to being pushed towards a _centralized_, single point of failure, javascript-heavy, bulky web frontend. I understand that Discourse has functionality for interacting via email, but that's fundamentally not its purpose, and it's apparent that email users are second class citizens. With mailing lists, most messages are sent as plain text email that displays properly in my TUI email client, and they're easy to quote and splice to reply to the parts in which I'm interested. I can both create new threads AND reply to existing ones.
It covers many different things at once, yet also drives us towards more scattered communications. Our infamous mega-threads are not really effective for getting to community consensus, and tend to bring out the worst in us.
How is this inherent to email? IMO, this correlates two unrelated things. I don't think it's fair to blame people's poor interpersonal communication on email.
I know this is a big change, but, hear me out…
I'll do my best, but I'll warn that I'm resistant change and am content with current mailing lists. If anything, I think we need to provide Fedora Infra more resources to properly maintain and keep up to date our Mailman instance.
We’re missing people
A Mastodon post from long-time Fedora contributor Major Hayden got me thinking:
How do people make so much time available for mailing list discourse?
Once I ensure my team has the technical guidance they need and I work through the tasks of work that I owe other people, I take a look at the mailing list and say: "Oh my gosh, what the heck happened here?" Then the discussion goes further off the rails while I'm typing out a reply and my reply is no longer relevant.
Again, I acknowledge this problem, but you've failed to demonstrate how this is inherent to mailing lists.
And… some people aren’t here because — in contrast with our “Friends” foundation, it isn’t always a nice place to be (and mailing lists don’t provide many tools for moderation, except the big hammer of outright bans).
Well, you can delete messages from the ML archives in the same way you can delete forum comments, but people who sign up for notifications will still receive a copy.
And that’s just one example. Take a look back at any mega-thread, and you’ll find similar — and worse. When things get heated, the only way to intervene is by adding more. There are often long subthreads of two people going back and forth on tangents. Then, other conversation branches duplicate that, or refer across. Classical email tools don’t actually handle this kind of thing very well at all. In my experience, it only really works if you keep up with the conversation in almost real time, which has its own problems even when that’s possible.
My email client has much better support for threading than Discourse does. I can easily toggle between threaded mode and flattened latest first mode. This is easier to navigate than Discourse's flat list. Email allows you to chose the best experience for you from a multitude of clients.
We’re scattered in actual practice
Devel list may be the center, but we have _hundreds_ of Fedora mailing lists. A dozen or so are reasonably active (Test, Legal, ARM…) but most are inactive or dead. Some are just meeting reminders over and over — for meetings that aren’t even active anymore. It’s easy to make but hard to _unmake_ a mailing list.
Yeah, I think the mailing lists could use more organization and cleanup.
For lists that are active, the split is confusing — when should something be on the packaging list rather than devel? What happens when something is related to both Cloud and Server, or Workstation and KDE? One can post to both lists, but if someone replies and isn’t subscribed to both, the conversation gets split.
Yeah, this can be confusing. I think the solution is organization and cleanup, not completely shutting down mailing lists.
With “devel” as the main list, conversations about marketing, design, events, and so on don’t really have a central place. (The Mindshare list never really caught on.) That makes these important activities feel even more disconnected and secondary in status — and they shouldn’t be.
Many groups have actually moved away from lists to using tickets for team conversations — both those non-engineering functions and development. Design Team has a mailing list, but mostly for announcements: the work happens in tickets. Workstation largely uses their Pagure tracker. And CoreOS conversations happen almost entirely in tickets on Github.
Gitlab being added to the mix definitely hasn't helped with the fragmentation. I think teams should be encouraged to standardize on tooling, but at the end of the day, teams will decide what works best for them. Freedom is one of the Four Foundations.
Tickets are made for tracking specific, actionable tasks, and that kind of tracking is part of why teams use them over mailing lists — but Fedora teams use them for open development conversations too.
I think having discussion split out from structured planning/ticket tracking is a good thing. The former is well suited for a mailing list or forum and the latter for issue trackers.
I think that’s largely a symptom of mailing lists not being enough for what we need. The trackers have media support, editing for typos or updates, reactions for simple agreement, tagging people, and granular subscriptions. They are effectively “off-label use” mini-forums that teams can quietly move to using without the sort of conversation I expect this message to generate.
Yes, those features are useful for some teams and are well suited for certain types of discussion, but I think the well established devel list has served us well and meets its purpose.
The set of remaining regular participants on this list is naturally biased towards those for whom it is working just fine. But, core Fedora development discussion can’t be limited to that ever-shrinking group. Consider who isn’t here. The problems are real, and the trend isn’t in a good direction.
What evidence shows that the group is ever shrinking? I often see Self Introduction posts and new people interacting with project. I suppose that whether they continue interacting afterwards is another question.
Devel List is too many different things!
We use it for Change discussion (resulting in those not-actually-so-great big threads).
We use it for introductions and onboarding — we’re usually pretty good at that, actually (but it adds to the overall load of following the list). We’re not very consistent, though.
We discuss packaging: guidelines, help on different topics, coordination on specific work. There’s unclear overlap with the packaging list, though, which is a bit confusing.
Yeah, I wonder if it makes sense to sunset the packaging list in favor of devel.
We talk about higher-level Fedora OS development topics that don’t fit anywhere else. For example, this on installer environment size: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org/... Sometimes, these are FESCo topics (which FESCo heroically struggles to keep here rather than on Pagure tickets…).
This became a problem for the incredibly controversial Frame Pointers change, but I'm not sure that's a general theme
Sometimes, people talk about a particular interest in the hopes of forming a Special Interest Group. Often that does result in a new SIG launching — sometimes with a brand-new mailing list which then ends up getting only a few posts ( and a whole lot of spam years later). But this really suffers from the survivorship problem: many people who might be interested will never notice.
We have release process announcements — mostly to devel-announce, but then occasionally replies and discussion here. Blocker bugs and other test-list and QA topics are cross-posted, as part of the release cycle.
And then there are a lot of robo-messages: reports, reminders, etc. These are really valuable to a few people, but add a lot more to wade through for everyone else trying to keep up.
There are certainly others, as well.
In short… there probably shouldn’t just be one thing. But, the cross-posting problem makes it hard to split up as a mailing list.
I think the devel list makes sense as a catch all, general development discussion place. People are free to filter out the threads that aren't relevant to them. Specific tasks that develop out of those discussions can be delegated to the respective SIG/WG/committee task tracker. The unified devel list allows cross-SIG communication and high level planning that I think we'd miss with fractured forum categories.
Enter Discourse
If you’ve talked with me about anything related to any of this in the past ten years, you probably already know that I like Discourse. It’s good software made by cool people. And, it’s entirely open source, with a SaaS business model but with real, usable releases. (No open core, no “open source theoretically but good luck”.)
I definitely appreciate that it's fully open source. This is more than I can say about other tools we're trying to move towards such as Gitlab.
And, we have it in production in Fedora already, at https://discussion.fedoraproject.org, with categories for announcements, user help, project discussion, and social conversation — as well as special categories for dedicated workflows.
Thank you for the work you put in to organizing and setting this up.
In Project Discussion, each different Fedora team can have its own tag, and you can subscribe to those that you’re interested in. Cross-posting is easy: tag a post with multiple teams.
Are there measures in place to avoid tag sprawl that makes it difficult to keep an eye on everything and subscribe?
Topics can be renamed, if needed, or split out into side-conversations. The long tangents from these conversations can actually be interesting on their own without distracting from the main topic. Moderation tools allow us to handle posts that are outside of expected Fedora contributor behavior, with varying levels of action as appropriate.
That is nice side effect of having a centralized platform. Of course, it can't solve problematic behavior, but it's nice to know that there's mitigation in place. It's just a question of whether these tools are used effectively and fairly and not excessively. I think I trust y'all to do that :). Still, I'd hope it's possible to improve moderation and encourage healthy dialogue without kludges.
You can use markdown formatting, including images (with easy addition of alt text for people for whom images are a barrier). You can edit your posts to fix typos or correct mistakes. There are polls and lots of other useful features.
These are definitely nice to have. I will note that these are incompatible with plain text email notifications which many here appreciate.
And, you can interact with it all by email. That interface isn’t perfect, but it’s way better than any other forum software I’ve seen. (I’ve written a guide: https://discussion.fedoraproject.org/t/25960) At the same time, your individual email address is hidden, so it won’t be a spam magnet (or a target for off-list harassment, a problem we unfortunately have with devel list).
Yeah, I'm not sure the email interface is really comparable to the ML experience. You can't post directly via email without the whole moderation queue thing mentioned in your linked post. Email interaction seems to be a second class citizen. I'm not sure what the point in moving off the mailing list is if you keep suggesting ways that people can make forum software behave more like the mailing list that we already have.
That said, it is web-first software. (Or mobile, if that’s your thing.) That requires some adjustment, I know. I hope opening up a Fedora Discussion tab – or keeping one open — becomes an easy habit.
I wouldn't like that much. The only tab I always have open is Matrix, and I'd prefer not to have to have more always open browser tabs. I already have an email client.
Not just Fedora
There’s a big trend towards Discourse in open source projects overall. Python and Gnome have both migrated entirely from their mailing lists. Ansible is working on it.
Ansible's mailing lists have been mostly announcements for a while. The plan is to move discussions and announcements away from various Github issue trackers and Github Discussions in various organizations into a central place. Fedora's level of mailing list usage and participation is significantly larger.
Plus, there’s Rust, Kubernetes, Nextcloud, Flathub, Grafana, Home Assistant, KDE, and I’m sure many others.
I'd argue that this is a type of fragmentation. Each has its own site that users have to separately sign up for and follow.
Concrete proposal
While I think it makes sense for new teams to consider adopting the forums instead of mailing lists, I am not convinced that we should shut down devel@ and alienate the _current_ devs who seem to favor the current approach.
Thanks for listening, Maxwell
-- Maxwell G (@gotmax23) Pronouns: He/They
On Fri, Apr 21, 2023 at 4:05 PM Maxwell G maxwell@gtmx.me wrote:
What evidence shows that the group is ever shrinking? I often see Self Introduction posts and new people interacting with project. I suppose that whether they continue interacting afterwards is another question.
I'm glad you asked. Earlier this week I decided to avoid doing other work by putting together some quick charts of devel list participation. Here's the number of unique participants per month from 2004–2022: https://bcotton.fedorapeople.org/images/devel-participation-monthly.png
And for a less-noisy version, the median of the monthly numbers per year: https://bcotton.fedorapeople.org/images/devel-participation-mean.png
There are a lot of questions left unanswered by this quick analysis, but there's a clear trend in fewer participants over time. In fact, last month had the second smallest participant count (tied with October 2022). Of course, these charts don't show _why_, but they do support the assertion that folks are dropping out of the conversation faster than others are joining.
On Fri, Apr 21, 2023 at 3:39 PM Ben Cotton bcotton@redhat.com wrote:
On Fri, Apr 21, 2023 at 4:05 PM Maxwell G maxwell@gtmx.me wrote:
What evidence shows that the group is ever shrinking? I often see Self Introduction posts and new people interacting with project. I suppose that whether they continue interacting afterwards is another question.
I'm glad you asked. Earlier this week I decided to avoid doing other work by putting together some quick charts of devel list participation. Here's the number of unique participants per month from 2004–2022: https://bcotton.fedorapeople.org/images/devel-participation-monthly.png
And for a less-noisy version, the median of the monthly numbers per year: https://bcotton.fedorapeople.org/images/devel-participation-mean.png
And on discourse you could have pasted these pictures directly in without having to upload them somewhere... ;)
There are a lot of questions left unanswered by this quick analysis, but there's a clear trend in fewer participants over time. In fact, last month had the second smallest participant count (tied with October 2022). Of course, these charts don't show _why_, but they do support the assertion that folks are dropping out of the conversation faster than others are joining.
I think this is especially concerning because Fedora seems to be more popular than ever. If we had a way to measure that and include it in the graph (devel participants as a % of users?) I have a feeling the slope would be much more negative.
Not that it proves Discourse would change that, but still...
Thanks, Richard
On Fri, Apr 21, 2023 at 04:38:28PM -0400, Ben Cotton wrote:
There are a lot of questions left unanswered by this quick analysis, but there's a clear trend in fewer participants over time. In fact, last month had the second smallest participant count (tied with October 2022). Of course, these charts don't show _why_, but they do support the assertion that folks are dropping out of the conversation faster than others are joining.
One obvious question: How does "sending mail to -devel" correlate with any other "contributing to Fedora" metric? Eg numbers of folks with Fedora accounts, number of packagers (or active packagers), git commiters/commits, and so forth? Or even numbers of installations, which to this way we can only kinda hand-wave about?
(I wonder how that also correlates to the likes of Debian?)
We also might be a victim of our own success here; things generally _work_ quite well, so there's not as much to go on about as there once was, and our various upstreams are increasingly driving our headliner features rather than the other way around.
So people simply don't _need_ to participate in "developing Fedora" like they once did to ensure their interests were represented, and instead get to focus more fully on the stuff they build on top. That's a good thing, I think, but it also takes away what was probably the primary participation funnel.
- Solomon
On 4/21/23 23:38, Ben Cotton wrote:
On Fri, Apr 21, 2023 at 4:05 PM Maxwell G maxwell@gtmx.me wrote:
What evidence shows that the group is ever shrinking? I often see Self Introduction posts and new people interacting with project. I suppose that whether they continue interacting afterwards is another question.
I'm glad you asked. Earlier this week I decided to avoid doing other work by putting together some quick charts of devel list participation. Here's the number of unique participants per month from 2004–2022: https://bcotton.fedorapeople.org/images/devel-participation-monthly.png
And for a less-noisy version, the median of the monthly numbers per year: https://bcotton.fedorapeople.org/images/devel-participation-mean.png
There are a lot of questions left unanswered by this quick analysis, but there's a clear trend in fewer participants over time. In fact, last month had the second smallest participant count (tied with October 2022). Of course, these charts don't show _why_, but they do support the assertion that folks are dropping out of the conversation faster than others are joining.
Ben, Thanks, this is helpful. Can you make the scripts/programs you used for these available to allow for a more detailed analysis?
On Sat, Apr 22, 2023 at 1:24 AM Benson Muite benson_muite@emailplus.org wrote:
Thanks, this is helpful. Can you make the scripts/programs you used for these available to allow for a more detailed analysis?
No, because I literally went to each page of the archives and copied the numbers into a spreadsheet. If you want to do a more in-depth analysis, you can download the mbox archive file.
On Sat, Apr 22, 2023 at 12:28 PM Zbigniew Jędrzejewski-Szmek zbyszek@in.waw.pl wrote:
Could we have the same graph for discourse (and Fedora telegram and Fedora matrix)? It'd be interesting to see what percentage of active communicating users are active on the mailing list.
I can't provide that, but someone could do that analysis. The hard part would be mapping the email addresses to Fedora accounts, especially as those may have changed over the years (and there are theoretically people on this list who don't have a Fedora account at all).
That speaks to some of the questions that come from the quick graphs I did: how many mailing list threads are of the zero-engagement announcement variety, versus active discussion? And what's the distribution of threads for users? Do most people reply to one or two threads and a handful reply to dozens?
-- Ben Cotton He / Him / His Fedora Program Manager Red Hat TZ=America/Indiana/Indianapolis
On Fri, Apr 21, 2023 at 04:38:28PM -0400, Ben Cotton wrote:
On Fri, Apr 21, 2023 at 4:05 PM Maxwell G maxwell@gtmx.me wrote:
What evidence shows that the group is ever shrinking? I often see Self Introduction posts and new people interacting with project. I suppose that whether they continue interacting afterwards is another question.
I'm glad you asked. Earlier this week I decided to avoid doing other work by putting together some quick charts of devel list participation. Here's the number of unique participants per month from 2004–2022: https://bcotton.fedorapeople.org/images/devel-participation-monthly.png
And for a less-noisy version, the median of the monthly numbers per year: https://bcotton.fedorapeople.org/images/devel-participation-mean.png
Could we have the same graph for discourse (and Fedora telegram and Fedora matrix)? It'd be interesting to see what percentage of active communicating users are active on the mailing list.
Zbyszek
On Sat, Apr 22, 2023 at 04:28:16PM +0000, Zbigniew Jędrzejewski-Szmek wrote:
Could we have the same graph for discourse (and Fedora telegram and Fedora matrix)? It'd be interesting to see what percentage of active communicating users are active on the mailing list.
I'm not sure how to get it from Matrix. We can get it from Discourse with some SQL queries. https://discourse.org/plugins/data-explorer.html
Once we have such a query defined, I can make it available for anyone (or certain groups of people) to run. (It's not generally available because you can query _a lot_.)
On Wed, Apr 26, 2023 at 10:51:33AM -0400, Matthew Miller wrote:
On Sat, Apr 22, 2023 at 04:28:16PM +0000, Zbigniew Jędrzejewski-Szmek wrote:
Could we have the same graph for discourse (and Fedora telegram and Fedora matrix)? It'd be interesting to see what percentage of active communicating users are active on the mailing list.
I'm not sure how to get it from Matrix. We can get it from Discourse with some SQL queries. https://discourse.org/plugins/data-explorer.html
Once we have such a query defined, I can make it available for anyone (or certain groups of people) to run. (It's not generally available because you can query _a lot_.)
The matrix side is a lot more muddy.
I can easily get via it's api number of users with the fedora.im homeserver. (1581).
But that is only users using fedora.im / chat.fedoraproject.org. There's all the other federated matrix users using whatever homeserver.
I can get events in rooms, but... we are bridging almost all our rooms to irc, so all the irc users events show up too on the other side of the bridge. I suppose some scripting could get all events and parse them for fedora.im/librea.chat/other? Anyone should be able to run such a script that can/is joined to the rooms.
From just my perception, The #fedora channel has had a lot more matrix users than irc users of late, and it's definitely increased since matrix was bridged in.
kevin
On 4/21/23 23:38, Ben Cotton wrote:
On Fri, Apr 21, 2023 at 4:05 PM Maxwell G maxwell@gtmx.me wrote:
What evidence shows that the group is ever shrinking? I often see Self Introduction posts and new people interacting with project. I suppose that whether they continue interacting afterwards is another question.
I'm glad you asked. Earlier this week I decided to avoid doing other work by putting together some quick charts of devel list participation. Here's the number of unique participants per month from 2004–2022: https://bcotton.fedorapeople.org/images/devel-participation-monthly.png
And for a less-noisy version, the median of the monthly numbers per year: https://bcotton.fedorapeople.org/images/devel-participation-mean.png
There are a lot of questions left unanswered by this quick analysis, but there's a clear trend in fewer participants over time. In fact, last month had the second smallest participant count (tied with October 2022). Of course, these charts don't show _why_, but they do support the assertion that folks are dropping out of the conversation faster than others are joining.
There are fewer participants than in the early heydays, but to me the overall participation figures look remarkably stable over the last ten years or so. Certainly not any "we fell off the cliff" type fall I expected to see based on the discussion.
- Panu -
On 4/21/23 23:04, Maxwell G wrote:
Hi Matthew,
On Thu Apr 20, 2023 at 17:20 -0400, Matthew Miller wrote
As it is, devel list is too much for many people to follow — people we’d like to have around.
I disagree. I find mailing lists much easier to deal with. They're all in one place, they're synced between my devices, and I can use whatever client I like. I can start writing a message on my phone, save it as a draft, and send it from my laptop. I can filter messages to my heart's content. Everything is threaded and I can choose to ignore the threads I'm uninterested in.
It's easy to skim the list and I can do so on my terms, as opposed to being pushed towards a _centralized_, single point of failure, javascript-heavy, bulky web frontend. I understand that Discourse has functionality for interacting via email, but that's fundamentally not its purpose, and it's apparent that email users are second class citizens. With mailing lists, most messages are sent as plain text email that displays properly in my TUI email client, and they're easy to quote and splice to reply to the parts in which I'm interested. I can both create new threads AND reply to existing ones.
Agree with this.
We’re scattered in actual practice
Devel list may be the center, but we have _hundreds_ of Fedora mailing lists. A dozen or so are reasonably active (Test, Legal, ARM…) but most are inactive or dead. Some are just meeting reminders over and over — for meetings that aren’t even active anymore. It’s easy to make but hard to _unmake_ a mailing list.
Yeah, I think the mailing lists could use more organization and cleanup.
Closing inactive and low traffic lists and moving them to Discourse is fine. Cleanup is helpful. Associated ticket: https://pagure.io/fesco/issue/2982
And, you can interact with it all by email. That interface isn’t perfect, but it’s way better than any other forum software I’ve seen. (I’ve written a guide: https://discussion.fedoraproject.org/t/25960) At the same time, your individual email address is hidden, so it won’t be a spam magnet (or a target for off-list harassment, a problem we unfortunately have with devel list).
Yeah, I'm not sure the email interface is really comparable to the ML experience. You can't post directly via email without the whole moderation queue thing mentioned in your linked post. Email interaction seems to be a second class citizen. I'm not sure what the point in moving off the mailing list is if you keep suggesting ways that people can make forum software behave more like the mailing list that we already have.
That said, it is web-first software. (Or mobile, if that’s your thing.) That requires some adjustment, I know. I hope opening up a Fedora Discussion tab – or keeping one open — becomes an easy habit.
I wouldn't like that much. The only tab I always have open is Matrix, and I'd prefer not to have to have more always open browser tabs. I already have an email client.
The web browser Matrix interface uses a lot of javascript and is very heavy and slow. Neochat and Nheko are much more performant than the electron based client. Have not tried the terminal based clients for Matrix. Discourse does seem to have an API, but email interaction is definitely a second class citizen. There is no lightweight performant text interface client for Discourse - it has not been designed for the power user.
Concrete proposal
While I think it makes sense for new teams to consider adopting the forums instead of mailing lists, I am not convinced that we should shut down devel@ and alienate the _current_ devs who seem to favor the current approach.
Much agreed. If it is challenging for people to join the development list, making some kind of email first bridge to Discourse or some other forum software is more reasonable. Not everything is meant to be web first. Some fragmentation in the Fedora ecosystem is good, it allows us to try new things. The bloated web is the reason for new protocols such as Gemini.
Thanks for listening, Maxwell
* Matthew Miller:
Big threads are … bad, actually
When we have something to talk about, it tends to explode into a big thread. The thing in January with FESCo’s frame pointers decision (https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org/...) is a good example of things going badly.
Most of the conversation was under the subject “Schedule for Tuesday's FESCo Meeting (2023-01-03)”, because everything started as a reply to that. That’s pretty easy to overlook. It’s possible for replies to change the subject when replying, but that can’t be done retroactively, and then isn’t consistent (and it breaks threading in Gmail, too).
Then, things got rather hostile, making it hard to have a reasonable conversation about the issues (both technical and procedural). And then, things went in circles without adding anything new.
This could have all gone a lot better.
You brought up fairly concrete examples, but I don't see how the platform matters in those cases. Someone would need to shut down or delete (parts of) conversations more or less arbitrarily, even if they are not actively harmful as such, simply because they create too much work for others to follow.
Surely that's not something you are willing to do?
For lists that are active, the split is confusing — when should something be on the packaging list rather than devel? What happens when something is related to both Cloud and Server, or Workstation and KDE? One can post to both lists, but if someone replies and isn’t subscribed to both, the conversation gets split.
Do Fedora mailing lists reject mail from non-members, and redirect follow-ups?
If they do that, that's the first thing you could fix. Then such discussions no longer would get split.
Not just Fedora
There’s a big trend towards Discourse in open source projects overall. Python and Gnome have both migrated entirely from their mailing lists. Ansible is working on it. Plus, there’s Rust, Kubernetes, Nextcloud, Flathub, Grafana, Home Assistant, KDE, and I’m sure many others.
All these instances are isolated. It's not possible to cross-post. This makes cross-project collaboration increasingly difficult because instead of Cc:ing another list, perhaps with a summary of the discussion so far, you have to create a completely new topic somewhere else, and then someone has to copy over summaries manually.
(If I recall correctly, that currently doesn't work well with Fedora lists because they reject posts from non-members.)
Personally, I have accounts on many, many Discourse instances, but I don't think there is a single one I read somewhat regularly. I find the mailing list mode and the notifications rather unpredictable. Maybe an alternative client could help (nndiscourse?), but as far as I understand it, there's no real API, so that's kind of hard?
First, I’d like to move the Changes discussion. They will still be posted to devel-announce, but responses directed to Project Discussion in a new #changes tag. Ben tells me that this is a FESCo decision, which seems reasonable.
There's already a Fesco ticket for each change, why not use that for discussion?
Other teams who want to keep mailing lists can, but I’d like to move those too, and eventually I think we’ll want to shut them down too — or perhaps convert them to announcement lists.
What about scm-commits?
Thanks, Florian
On Fri, 2023-04-21 at 23:20 +0200, Florian Weimer wrote:
For lists that are active, the split is confusing — when should something be on the packaging list rather than devel? What happens when something is related to both Cloud and Server, or Workstation and KDE? One can post to both lists, but if someone replies and isn’t subscribed to both, the conversation gets split.
Do Fedora mailing lists reject mail from non-members, and redirect follow-ups?
Many lists *hold* mail from non-members, because mailing lists get tons of spam. So the mail won't get through until an admin approves it. That might happen right away...or it might happen in two days, when the mail is no longer relevant. We can't really just let all mails from non- members through because...spam.
On Fri, Apr 21, 2023 at 02:30:45PM -0700, Adam Williamson wrote:
On Fri, 2023-04-21 at 23:20 +0200, Florian Weimer wrote:
For lists that are active, the split is confusing — when should something be on the packaging list rather than devel? What happens when something is related to both Cloud and Server, or Workstation and KDE? One can post to both lists, but if someone replies and isn’t subscribed to both, the conversation gets split.
Do Fedora mailing lists reject mail from non-members, and redirect follow-ups?
Many lists *hold* mail from non-members, because mailing lists get tons of spam. So the mail won't get through until an admin approves it. That might happen right away...or it might happen in two days, when the mail is no longer relevant. We can't really just let all mails from non- members through because...spam.
Right. I don't think we have many (or possibly any) lists that still hold email from non-members. The flood of spam is just too high for that for the last N years. So, almost all our lists are set to reject non member posts. :(
kevin
On Sat, 2023-04-22 at 10:37 -0700, Kevin Fenzi wrote:
On Fri, Apr 21, 2023 at 02:30:45PM -0700, Adam Williamson wrote:
On Fri, 2023-04-21 at 23:20 +0200, Florian Weimer wrote:
For lists that are active, the split is confusing — when should something be on the packaging list rather than devel? What happens when something is related to both Cloud and Server, or Workstation and KDE? One can post to both lists, but if someone replies and isn’t subscribed to both, the conversation gets split.
Do Fedora mailing lists reject mail from non-members, and redirect follow-ups?
Many lists *hold* mail from non-members, because mailing lists get tons of spam. So the mail won't get through until an admin approves it. That might happen right away...or it might happen in two days, when the mail is no longer relevant. We can't really just let all mails from non- members through because...spam.
Right. I don't think we have many (or possibly any) lists that still hold email from non-members. The flood of spam is just too high for that for the last N years. So, almost all our lists are set to reject non member posts. :(
ah, I hadn't noticed that change :/ I could've sworn I still sometimes get hold notices when I send meeting announcements to lists I'm not subscribed to...
On Sat, 2023-04-22 at 12:16 -0700, Adam Williamson wrote:
On Sat, 2023-04-22 at 10:37 -0700, Kevin Fenzi wrote:
On Fri, Apr 21, 2023 at 02:30:45PM -0700, Adam Williamson wrote:
On Fri, 2023-04-21 at 23:20 +0200, Florian Weimer wrote:
For lists that are active, the split is confusing — when should something be on the packaging list rather than devel? What happens when something is related to both Cloud and Server, or Workstation and KDE? One can post to both lists, but if someone replies and isn’t subscribed to both, the conversation gets split.
Do Fedora mailing lists reject mail from non-members, and redirect follow-ups?
Many lists *hold* mail from non-members, because mailing lists get tons of spam. So the mail won't get through until an admin approves it. That might happen right away...or it might happen in two days, when the mail is no longer relevant. We can't really just let all mails from non- members through because...spam.
Right. I don't think we have many (or possibly any) lists that still hold email from non-members. The flood of spam is just too high for that for the last N years. So, almost all our lists are set to reject non member posts. :(
ah, I hadn't noticed that change :/ I could've sworn I still sometimes get hold notices when I send meeting announcements to lists I'm not subscribed to...
In theory we could make it simpler by sending back a message that requires just a click to subscribe/authorize the email by a real user, if they intend to do so, on their first email to a mailing list. We could also allow posting to other mailing lists if the email address is subscribed to any other list.
I realize this would require working on mailman and that is probably something we do not want to spend time on ...
After all you have to subscribe to discourse as well to be able to post ... so there is no huge difference here.
On Mon, 24 Apr 2023 at 09:38, Simo Sorce simo@redhat.com wrote:
On Sat, 2023-04-22 at 12:16 -0700, Adam Williamson wrote:
On Sat, 2023-04-22 at 10:37 -0700, Kevin Fenzi wrote:
On Fri, Apr 21, 2023 at 02:30:45PM -0700, Adam Williamson wrote:
On Fri, 2023-04-21 at 23:20 +0200, Florian Weimer wrote:
For lists that are active, the split is confusing — when should something be on the packaging list rather than devel? What
happens when
something is related to both Cloud and Server, or Workstation
and KDE?
One can post to both lists, but if someone replies and isn’t
subscribed
to both, the conversation gets split.
Do Fedora mailing lists reject mail from non-members, and redirect follow-ups?
Many lists *hold* mail from non-members, because mailing lists get
tons
of spam. So the mail won't get through until an admin approves it.
That
might happen right away...or it might happen in two days, when the
is no longer relevant. We can't really just let all mails from non- members through because...spam.
Right. I don't think we have many (or possibly any) lists that still hold email from non-members. The flood of spam is just too high for
that
for the last N years. So, almost all our lists are set to reject non member posts. :(
ah, I hadn't noticed that change :/ I could've sworn I still sometimes get hold notices when I send meeting announcements to lists I'm not subscribed to...
In theory we could make it simpler by sending back a message that requires just a click to subscribe/authorize the email by a real user, if they intend to do so, on their first email to a mailing list. We could also allow posting to other mailing lists if the email address is subscribed to any other list.
We had this running for a bit where we would send back emails saying this is held. 99% of the emails would then go sit in queue on bastion slowing down regular deliveries. We are talking hundreds of emails a day on 'good' days and tens of thousands on 'bad' days. Trying to deal with this is a full time job that no one is paid (and my volunteer time is limited) to do. Trying to fix in better ways is usually a massive project because you need to think out the total email flow plan and needs. Email is no longer the old 'set up a mail server and let it live'. It is 'why do we have 10,000 queued emails today?' 'why aren't redhat.com emails getting delivered today?' 'oh look 2 new DNS features to 'deal' with SPAM', 'oh spamassassin needs new setup and fixes', 'why is email stuck here?' 'why is X sending email to google.com but we are getting errors from gandhi.net or protonmail.com'.
The software which runs mailing lists is also much more complicated than it was 20 years ago. You need to deal with backend databases, caching web servers, internal search engines, message tooling, spam deletion, account acceptance, etc. That takes constant learning and dedicated 'brain' space for admins to keep it working.
In order to keep email working, it takes dedicated and hard work and decisions to make it happen.
I realize this would require working on mailman and that is probably
something we do not want to spend time on ...
After all you have to subscribe to discourse as well to be able to post ... so there is no huge difference here.
-- Simo Sorce RHEL Crypto Team Red Hat, Inc
devel mailing list -- devel@lists.fedoraproject.org To unsubscribe send an email to devel-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
* Kevin Fenzi:
Right. I don't think we have many (or possibly any) lists that still hold email from non-members. The flood of spam is just too high for that for the last N years. So, almost all our lists are set to reject non member posts. :(
That's really unfortunate because it makes it harder to collaborate, not just for newcomers. You really need to be in control of your email setup and make sure you match addresses. It also makes mail-based cross-distribution collaboration very difficult. Other mailing list operators do not seem to have to do this, while maintaining acceptable spam levels.
Anyway, the present state of Fedora mailing lists sort of addresses my concerns about moving Fedora into a Discourse silo because Fedora seems to have removed itself from the collaborative email sphere a while ago anyway.
Thanks, Florian
On 4/21/23 23:20, Florian Weimer wrote:
- Matthew Miller:
Big threads are … bad, actually
When we have something to talk about, it tends to explode into a big thread. The thing in January with FESCo’s frame pointers decision (https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org/...) is a good example of things going badly.
Most of the conversation was under the subject “Schedule for Tuesday's FESCo Meeting (2023-01-03)”, because everything started as a reply to that. That’s pretty easy to overlook. It’s possible for replies to change the subject when replying, but that can’t be done retroactively, and then isn’t consistent (and it breaks threading in Gmail, too).
Then, things got rather hostile, making it hard to have a reasonable conversation about the issues (both technical and procedural). And then, things went in circles without adding anything new.
This could have all gone a lot better.
You brought up fairly concrete examples, but I don't see how the platform matters in those cases. Someone would need to shut down or delete (parts of) conversations more or less arbitrarily, even if they are not actively harmful as such, simply because they create too much work for others to follow.
Surely that's not something you are willing to do?
For lists that are active, the split is confusing — when should something be on the packaging list rather than devel? What happens when something is related to both Cloud and Server, or Workstation and KDE? One can post to both lists, but if someone replies and isn’t subscribed to both, the conversation gets split.
Do Fedora mailing lists reject mail from non-members, and redirect follow-ups?
If they do that, that's the first thing you could fix. Then such discussions no longer would get split.
Not just Fedora
There’s a big trend towards Discourse in open source projects overall. Python and Gnome have both migrated entirely from their mailing lists. Ansible is working on it. Plus, there’s Rust, Kubernetes, Nextcloud, Flathub, Grafana, Home Assistant, KDE, and I’m sure many others.
All these instances are isolated. It's not possible to cross-post. This makes cross-project collaboration increasingly difficult because instead of Cc:ing another list, perhaps with a summary of the discussion so far, you have to create a completely new topic somewhere else, and then someone has to copy over summaries manually.
(If I recall correctly, that currently doesn't work well with Fedora lists because they reject posts from non-members.)
Personally, I have accounts on many, many Discourse instances, but I don't think there is a single one I read somewhat regularly. I find the mailing list mode and the notifications rather unpredictable. Maybe an alternative client could help (nndiscourse?), but as far as I understand it, there's no real API, so that's kind of hard?
I could find an API docs, and I could retreive posts.json from our Fedora instance
So the question is, what is a "real API" that you would consider OK? For a post, there is create, update, delete, retrieve and even list the latest. IMHO that's as CRUD as one can be :D.
So, there is even a REST API? At least retrieval can go unauthenticated...
Regards, jarek
* Jarek Prokop:
Personally, I have accounts on many, many Discourse instances, but I don't think there is a single one I read somewhat regularly. I find the mailing list mode and the notifications rather unpredictable. Maybe an alternative client could help (nndiscourse?), but as far as I understand it, there's no real API, so that's kind of hard?
I could find an API docs, and I could retreive posts.json from our Fedora instance
So the question is, what is a "real API" that you would consider OK?
There has to be a login procedure, and it needs to be geared towards alternative clients. We currently have only this:
| To become authenticated you will need to create an API Key from the | admin panel.
So its seems to be restricted to admin-approved integrations, and is not intended for writing clients.
Thanks, Florian
On 25/04/2023 09:40, Florian Weimer wrote:
- Jarek Prokop:
Personally, I have accounts on many, many Discourse instances, but I don't think there is a single one I read somewhat regularly. I find the mailing list mode and the notifications rather unpredictable. Maybe an alternative client could help (nndiscourse?), but as far as I understand it, there's no real API, so that's kind of hard?
I could find an API docs, and I could retreive posts.json from our Fedora instance
So the question is, what is a "real API" that you would consider OK?
There has to be a login procedure, and it needs to be geared towards alternative clients. We currently have only this:
| To become authenticated you will need to create an API Key from the | admin panel.
So its seems to be restricted to admin-approved integrations, and is not intended for writing clients.
It is possible to allow users to generate their own API keys without any admin involvement but there's no direct web interface to create one - the client has to make a call to /user-api-key/new with certain parameters - full details are here:
https://meta.discourse.org/t/user-api-keys-specification/48536
Tom
On Fri, 21 Apr 2023 at 17:20, Florian Weimer fweimer@redhat.com wrote:
For lists that are active, the split is confusing — when should something be on the packaging list rather than devel? What happens when something is related to both Cloud and Server, or Workstation and KDE? One can post to both lists, but if someone replies and isn’t subscribed to both, the conversation gets split.
Do Fedora mailing lists reject mail from non-members, and redirect follow-ups?
Yes we have to. Most of the email coming to any of the fedora mailing lists is spam email from non-subscribers. A good portion of it is 'smart' SPAM where it replies to a specific email with headers to make it pop into an existing thread. Fiddling with various spam controls to better handle that has continually caused important developers emails to start being marked as SPAM.
The first thing we did was to have non-member email moderated for the lists. This sounds great but the amount of queued email on many lists is in the order of 100k or more emails. The problem is that the version of mailman3 uses some sort of linked list to keep track of all those queued emails. Every email seems to get checked against that queue which then slows down the overall system. Last year we were getting hour long timeouts on mailman3, and I then spent about 3 man-weeks of volunteer time to go through only about 100 mailing lists to clean out the 100k queues on each of them. I stopped when the timeouts got down to a 'normal' amount but the amount of junk email on the many email lists is a lot. There are probably ways to do this directly in the postgres database, but my attempts required restores from backups due to 'differences between our beta setup and what is expected'.
Trying to upgrade mailman3 to a version which may be better has been,I think, a 5 year task of continual frustration. When I was in infrastructure, everyone always had about 10 other tasks of higher priority that HAD to be done to keep other parts of Fedora running. When I left infrastructure, that increased the tasks for the remaining people to 20. Hiring in a replacement just found more things which needed to be kept running so we have looked for volunteers for a while. Several attempts have been made by volunteers, but real life and the overall complexity of modern email kills it every time.
Running mailman3 is a nearly full time job to keep it working versus the lackadaisical mailman2 it replaced. Because it is trying to be both a webforum to catch that 'I don't want to use email' audience, a better archiver, and various other tooling, Things like system accounts, authentication, postgres databases, etc They are all needed to make it work.
Outside of that DNS has many new fields which need to be implemented or added to deal with slightly conflicting standards which cause various sites to not accept email if they aren't implemented in their version. New fields are added and changed regularly which require dealing with people complaining that their email is now marked as SPAM, they aren't getting the email anymore, or that we have lost email because the queues on our systems overflowed due to various people subscribing to the SCM mailing list but having a quota too small.
In any case, the issue is that there have not been for about 8 years to run this well. The task gets harder and harder over time due to complex DNS needs for email to work these days to just general time needed to clean up existing spam, deal with ham being marked as spam, etc. And when it comes down to 'does Infrastructure have time to keep builds, composes, and the 100 services running that are needed to do that' or 'does Infrastructure work on some part of an undocumented email system'.. the answer is always going to be get the daily builds out as developers complain a lot more about that than email.
It is time to explore other options. One of them is the proposal that Matthew and I guess the Council have come up with. It is using a resource which is paid for, has an open source background, and is willing to make some changes to better accommodate other workflows. If people want something else they are going to need to come up with a proposal which does not include using existing burned out resources to accomplish it.
* Stephen Smoogen:
It is time to explore other options. One of them is the proposal that Matthew and I guess the Council have come up with. It is using a resource which is paid for, has an open source background, and is willing to make some changes to better accommodate other workflows. If people want something else they are going to need to come up with a proposal which does not include using existing burned out resources to accomplish it.
Have you considered outsourcing email (list) operations instead?
Discourse needs to solve the same issues to keep email notifications working. Do we know how well they are doing? Discourse seems to have stopped offering mailing list mode in the Fedora instance. I wonder whether this is related.
Thanks, Florian
On Tue, Apr 25, 2023 at 10:49:03AM +0200, Florian Weimer wrote:
Have you considered outsourcing email (list) operations instead?
We looked into a hosted Pony Mail a while ago. But (leaving aside "I don't think that solves everything I want to address"), hosted solutions are difficult because if Red Hat is paying for it, there is an enterprise-contract process to go through, and generally a lot of vendor scrutiny and demands -- much of which is good and some of which I feel is overkill. It took over a year each to get Element (Matrix) and CDCK (the Discourse company) through the process -- and those companies had already dealt with enterprise contracts. I think it's unlikely we'll find a good open source option.
Discourse needs to solve the same issues to keep email notifications working. Do we know how well they are doing? Discourse seems to have stopped offering mailing list mode in the Fedora instance. I wonder whether this is related.
They're ahead of us, I think. I was recently talking to them about the possibility of using @fedoraproject.org sender / inbound addresses, and got an overview of their infrastructure and it's enough that it made me bookmark the whole thing for when I have more time. :)
I turned off the mailing list mode on purpose. There's a subthread somewhere here about it. I feel like it's a trap -- it does not really enable anything special, but instead is "subscribe to ALL LISTS". That's not the best option for most people -- even people who are looking for mailing-list behavior. (It might made sense for smaller forums where the entire forum is the equivalent of one list.)
If there's enough demand, we can put it back. As mentioned in the other subthread, if so I'll probably rename publicly-facing option to something more descriptive.
Once upon a time, Matthew Miller mattdm@fedoraproject.org said:
I turned off the mailing list mode on purpose.
So... this brings up something about this whole thread that I've avoided but think does matter: I understand that you are the Fedora Project Leader, but it seems like a lot of this is being done based on your personal decisions. I believe that you are doing what you think is best for the project (NOT questioning your motives, experience, etc.), but is all of this entirely in scope for the project leader to decide on their own?
You believe that the lists are outdated/blocking new contributors, but where is the evidence to support that (and that a web forum will be better), and who else agreed to such a change? You turned off mailing list mode because you believe it isn't good, but who else had input?
I'm just a little concerned that because you don't like mailing lists, Fedora must migrate away, and to something where you'll be making more individual decisions.
On Wed, Apr 26, 2023 at 10:20:11AM -0500, Chris Adams wrote:
So... this brings up something about this whole thread that I've avoided but think does matter: I understand that you are the Fedora Project Leader, but it seems like a lot of this is being done based on your personal decisions. I believe that you are doing what you think is best for the project (NOT questioning your motives, experience, etc.), but is all of this entirely in scope for the project leader to decide on their own?
I think there's two mixed things here.
For the forum itself, I've been acting as admin, and I admit I'm excited about it. But I haven't been making most decisions alone — I usually post in the public Site Help & Feedback forum for discussion first. (See these topics for some big examples: https://discussion.fedoraproject.org/t/considering-a-general-reorganization-... https://discussion.fedoraproject.org/t/considering-a-merge-of-ask-fedora-int...) ) But in the future, and probably not the very far future, we should have a formal team around this.
Second, there's the decision about what to do with devel list, or mailing lists overall. As outlined in the first post, I'm not making that decision alone. We've dicussed it in the Fedora Council, and we're having this discussion here, and the first step is the intended experiment with the Change process -- which is a FESCo decision.
You believe that the lists are outdated/blocking new contributors, but where is the evidence to support that (and that a web forum will be better), and who else agreed to such a change? You turned off mailing list mode because you believe it isn't good, but who else had input?
I feel like this is actually an example of the problem with so-called "mailing list mode". Out of context, that sounds like I have disabled a mode that makes Discourse behave more like a mailing list. (I mean, right?) But that's not what it does, as I've noted several places already here. So when upstream disabled the setting by default (https://meta.discourse.org/t/2-7-0-beta5-improved-invites-auto-tag-and-auto-...) it made sense to me to me to just follow that change.
I'm just a little concerned that because you don't like mailing lists, Fedora must migrate away, and to something where you'll be making more individual decisions.
Honestly, it's just plain not sustainable for me to keep doing that, whether or not it's good or bad. :)
I'm relatively new on mailing lists but just wanted to give my opinion.
With mailing lists one gets to choose the client to use, unlike with the web based forums where one is more or less limited to use a browser. The Discoure's mailing list functionality sounds interesting but how much it takes effort to configure and tune tags/categories to follow? Are the tags self explanatory or will there be brief info of which tag contains what?
I would consider myself still as newcomer and I didn't find self-introduction and joining to a mailing list a problem. What comes to participation, this is my second post, so I haven't been taking part of conversation very actively. Switching to Discourse would not make me post more (unless there is magically more topics which I could participate in).
Regards, Tomi Lähteenmäki
I would also be one of those people who would be much less engaged -- even disengage -- if everything moved to a website.
I have one thing to add that I don't think was covered:
Can we make contributing to the mailing list easier?
One thing we have done in a community I manage is *not* to require any sign up to the mailing list before posting. They are invited to simply email the list address. The flip side to this is obviously that "someone" (cough, me) has to filter out a lot of spam carefully. It's not too bad in that small community, but might be a lot more work in Fedora. But it ought to reduce the barrier to entry to "able to send an email" which is IMHO pretty low.
Rich.
On Sat, Apr 22, 2023 at 06:03:12PM +0100, Richard W.M. Jones wrote:
I would also be one of those people who would be much less engaged -- even disengage -- if everything moved to a website.
I have one thing to add that I don't think was covered:
Can we make contributing to the mailing list easier?
One thing we have done in a community I manage is *not* to require any sign up to the mailing list before posting. They are invited to simply email the list address. The flip side to this is obviously that "someone" (cough, me) has to filter out a lot of spam carefully. It's not too bad in that small community, but might be a lot more work in Fedora. But it ought to reduce the barrier to entry to "able to send an email" which is IMHO pretty low.
See smooge's reply here: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org/...
Zbyszek
On Sat, 22 Apr 2023 at 13:03, Richard W.M. Jones rjones@redhat.com wrote:
I would also be one of those people who would be much less engaged -- even disengage -- if everything moved to a website.
I have one thing to add that I don't think was covered:
Can we make contributing to the mailing list easier?
One thing we have done in a community I manage is *not* to require any sign up to the mailing list before posting. They are invited to simply email the list address. The flip side to this is obviously that "someone" (cough, me) has to filter out a lot of spam carefully. It's not too bad in that small community, but might be a lot more work in Fedora. But it ought to reduce the barrier to entry to "able to send an email" which is IMHO pretty low.
Fedora mailman gets a lot of emails. Generally we would see about 900 emails per day that would need to be dealt with per day for the devel list and 1600 emails per day for users. The lower end mailing lists see about 200 emails a day that come in and get dropped from non-subscribers. Processing those is not a fast issue with mailman3 because of database and schema issues. It generally takes 20 to 30 seconds to delete or accept a held email. Trying to do them in bulk halts ALL other transactions as it has to lock various tables. Some of these issues may be fixed in a newer version of mailman3, but we can't just update to any newer version because our database schema was a prerelease.
On Sat, Apr 22, 2023 at 06:03:12PM +0100, Richard W.M. Jones wrote:
I would also be one of those people who would be much less engaged -- even disengage -- if everything moved to a website.
I have one thing to add that I don't think was covered:
Can we make contributing to the mailing list easier?
One thing we have done in a community I manage is *not* to require any sign up to the mailing list before posting. They are invited to simply email the list address. The flip side to this is obviously that "someone" (cough, me) has to filter out a lot of spam carefully. It's not too bad in that small community, but might be a lot more work in Fedora. But it ought to reduce the barrier to entry to "able to send an email" which is IMHO pretty low.
We used to do this for a long time, but it became untenable a number of years ago. When you look at there's 10 posts to look thru for legit non members thats fine, when you look and there's 10,000...
:(
kevin
On Thu, Apr 20, 2023 at 05:20:37PM -0400, Matthew Miller wrote:
First, I’d like to move the Changes discussion. They will still be posted to devel-announce, but responses directed to Project Discussion in a new #changes tag. Ben tells me that this is a FESCo decision, which seems reasonable.
I think this is a great place to start. If it works, wonderful. If it doesn't work, then we can delay/abort/redesign any later steps.
Second, I think other FESCo-related conversations should move. I hope this will reduce the urge to have back-and-forth exchanges in the tickets. For the Fedora Council, I set up a bot which automatically creates a discussion topic when a ticket is filed, leaving the ticket just for votes and recording of outcome. FESCo could use something similar.
Could we instead have a poll which is open only to members in a FAS group and tally those votes? I guess this would be useful for other things too. (E.g. as a replacement for blocker-review [1]).
[1] https://pagure.io/fedora-qa/blocker-review
And finally… shut down the devel list itself. Perhaps at the end of 2023?
That seems way too early. Let's not plan for this until it is clear that we can do that without losing contributors.
We should also shut down all of the little lists that haven’t had anything but spam in the last two years. We could maybe do that sooner. We should stop creating new lists now — we can create new Discussion tags instead.
Ack. There's also https://pagure.io/fesco/issue/2982 on the same topic now.
For right now, though: let’s discuss — on the list!
Thank you for this detailed proposal. I think it was very clear and even and written in a way that led the discussion in good directions.
Zbyszek
On Sat, Apr 22, 2023 at 05:40:41PM +0000, Zbigniew Jędrzejewski-Szmek wrote:
I think this is a great place to start. If it works, wonderful. If it doesn't work, then we can delay/abort/redesign any later steps.
Exactly.
Second, I think other FESCo-related conversations should move. I hope this will reduce the urge to have back-and-forth exchanges in the tickets. For the Fedora Council, I set up a bot which automatically creates a discussion topic when a ticket is filed, leaving the ticket just for votes and recording of outcome. FESCo could use something similar.
Could we instead have a poll which is open only to members in a FAS group and tally those votes? I guess this would be useful for other things too. (E.g. as a replacement for blocker-review [1]).
Yes, that's possible. (The group sync stuff is almost in place -- infra SOP docs getting finalized.)
[1] https://pagure.io/fedora-qa/blocker-review
And finally… shut down the devel list itself. Perhaps at the end of 2023?
That seems way too early. Let's not plan for this until it is clear that we can do that without losing contributors.
Fair enough! I intend to only submit the Changes move to FESCo at this time (or, once this discussion has settled), and the rest can come as it does. I just wanted to also give a clear picture of what I'm aiming for right up front.
Il 20/04/23 23:20, Matthew Miller ha scritto:
It’s time to transform the Fedora devel list into something new
... Matthew Miller mattdm@fedoraproject.org Fedora Project Leader
I've tried to follow up all the replies this post has caused, but it takes time to read in a foreign language and there are simply too many posts to read them all, so I'll just go and give you some thoughts that may have already been written by someone else.
- I'm not enthusiast of being forced to open a web browser to be noticed about new posts. I'm aware there's a "mailing list" mode, but it's not clear if it's perfectly usable or would be sub-optimal. I think it needs to be tested.
- I'm also skeptical about having all mailing lists under one Discourse category (I suppose it will be "Project Discussions") and use tags to filter them. I think an high volume list such as devel should gone under it's own category. Maybe for other low volumes lists which are specific to groups or aspects can use the "tags" categorization.
- Another problem is that there are already too many tags available and the tag description is not visible in the post submit form. I think a new user would be confused what tag to use. I'm confused too, but that's probably just me...
- The Discourse Android app is no use. So, again, browser on mobile or mailing list mode.
Mattia
On Sun, Apr 23, 2023 at 05:23:34PM +0000, Mattia Verga via devel wrote:
- I'm also skeptical about having all mailing lists under one Discourse
category (I suppose it will be "Project Discussions") and use tags to filter them. I think an high volume list such as devel should gone under it's own category. Maybe for other low volumes lists which are specific to groups or aspects can use the "tags" categorization.
I think the several different functions of the list belong in different tags. What's here and what's on some other list (or already on a different platform altogether) is largely an accident of history.
If it does seem like there's a clear set of things which could be split to a different category, we have that option in the future. The category of a topic is just metadata, so chagning it is fast and doesn't change URLs or anything.
- Another problem is that there are already too many tags available and
the tag description is not visible in the post submit form. I think a new user would be confused what tag to use. I'm confused too, but that's probably just me...
That's useful feedback. We're working on making that better.
Hello Matthew, everyone,
The TL;DR version of my thoughts and suggestions on this topic: implement the improvements suggested so far as best as possible, so that the people who want to interact with Discourse through their mail reader can do so with little to no friction. Clearly define and document the new tags, what type of information should go where and who needs to subscribe to which tag. Redefine our information channels in a saner way, that reflects our consumption and use of that information. Document on the docs website how subscriptions, filtering etc. work in the Discourse UI for each use case. Avoid further community fragmentation by switching cold turkey once the organization of the site and the mailing features are on par with what people have come to expect from mailing lists. Turn on the requirement of adhesion to a fas group for people who can post on the new channels.
Longer version: Like most of the people who have chimed in so far, I too have strong opinions and feelings about such a transition. I joined the mailing list almost a decade ago and I have been following it for much longer. I am subscribed to a number of lists, I set up my filters ages ago and over time, as the project and my involvement with it changed, I adapted those filters with minimal effort. I can certainly sympathize with all the people who are far more active than I am and who interact with Fedora first and foremost through these lists, even though I don't go so far as to run an NNTP server or type my replies in vi (admittedly, that was fun for a while).
Since this transition is more or less a settled issue, as far as I can understand, I would like to make some suggestions as to how this could be as painless as possible for everyone. Before that though, I'd like to take a moment to point out something that has been very lightly touched upon, the nature of the medium (e-mail) and its non-technical aspects. I think most people are acquainted with the formalisms and the urgency/importance attributed to different means of communication. We don't feel the same way about a written letter we found in our mailbox, a phone call, an e-mail, an instant message, an SMS or a forum post. E-mail commands a certain "decorum" close to that of a written letter. Perhaps this is the reason why certain people find it daunting to engage with the project through the mailing lists. At the same time and while we're no strangers to heated exchanges and a far cry from literary correspondence, the etiquette dictated by the medium does create a safe space for its users. We all pretty much understand and adhere to the rules associated with this kind of written communication and have certain expectations when we use it and even more so given our relationship to Fedora. This is not the case with a forum and certainly not the case with all the people who will go on d.f.o. For this reason I strongly suggest that the "Project Discussion" category or whatever subcategory our particular subcommunity ends up on, is indeed walled off from the rest of the website. I do not need the noise, the rudeness and the abuse that I often found on ask.fedoraproject.org (which is now a part of d.f.o) to flood over these channels. The requirement on people belonging to a specific fas group before they can post should be on from day one and exceptions made only when they are warranted.
On Thu, Apr 20, 2023 at 11:21 PM Matthew Miller mattdm@fedoraproject.org wrote:
We’re scattered in actual practice
This is not necessarily or always a bad thing, nor is it a problem that can be solved by the transition to Discourse. I thought I had a fairly good grasp of most of Fedora's activities and communities until I saw the organizational chart posted by Marie Nordin a while back and I realized that about a third of Fedora was terra incognita to me. Then again, we cannot all afford the time nor do we have the skills and proclivities required to be involved with everything. The devel ML has been a good "firehose" of information for people who work on packaging and development and everything tangentially tied to these. It's only natural that the stream of information will get fragmented down the road, but we've mostly had a "bird's eye view" of the issues of interest to us through this channel and we could (usually) follow up on whatever place a particular discussion unfolded, be it bugzilla, a pagure ticket or whatever.
The "announce" ML could become an "announce" category for all things Fedora, which might be of interest to people from different subcommunities, while keeping it relatively low-traffic. E.g. voting for the default wallpaper for the next Fedora release is something that would fit that description, whereas a release party in Paraguay or the move to Python 3.12 would not.
If the mailing lists and Discourse are being used simultaneously, we will end up even more scattered than we are now. If the concerns voiced so far can be addressed and solutions implemented in Discourse in a way that does not disrupt people's work flows and does not pose a significant burden to switch (essentially reinventing Hyperkitty in the opposite direction), then I think it would do us good to switch immediately.
Devel List is too many different things!
This is a good opportunity to untangle the mess and redefine what type of information should go to which channel, keeping in mind the interests of the intended audience. It won't make any sense to split up devel discussions over 20 different tags to all of which all of us will be subscribed to again. Reports, reminders, notifications of blocker bugs, etc. are all things potentially interesting to the devel audience and easily filtered as long as they maintain a constant formatting. Can we have nested tags maybe?
While it might not make sense to have a 1-1 transition of mailing lists to tags, the tags should be well thought-out in advance; it should be made clear what goes where and who must subscribe to what. Clear, take-you-by-the-hand documentation on how these things should be set up on the Discourse interface is absolutely vital.
First, I’d like to move the Changes discussion. They will still be posted to devel-announce, but responses directed to Project Discussion in a new #changes tag.
Absolutely vital to every devel(-announce) subscriber, it would make for a great nested tag, if that were an option.
Second, I think other FESCo-related conversations should move. I hope this will reduce the urge to have back-and-forth exchanges in the tickets. For the Fedora Council, I set up a bot which automatically creates a discussion topic when a ticket is filed, leaving the ticket just for votes and recording of outcome. FESCo could use something similar.
How will the current devel audience keep track of these discussions?
Fourth, I’d like to update our documentation, process, and expectations for newcomers — say hello on Discussion (and Fedora Chat, if you like) rather than a mailing list. (I’d like to close the Fedora Join list at this point.)
Newcomers should be connected to the people in the communities which they intend to join. Graphics artists should introduce themselves in #fedora-design (made-up name, don't know if it exists), packagers in #devel, people wishing to do PR work in #marketing and so on. If the introductions are done elsewhere, no one (relevant) will ever meet them, at least those on devel. (We have a Fedora Join ML???)
Fifth, all packaging-related discussions (including the separate packaging mailing list). We already have a #package-maintainers tag with some existing discussion.
As long as it's clear and documented, splitting that off from devel would be ok, but are the audiences different?
Sixth, automated posts, as much as we can. These should go to dedicated Workflow categories, where people who want can watch them but where they won’t overwhelm human interactions. People who want can watch them, and it’s easy to quote-reply into a new linked topic in the Project Discussions category.
For many people, this type of information is important some of the time. It would be preferable to have it available and easy to filter out and/or delete, rather than keeping it out of sight and out of mind.
Matthew Miller wrote:
I propose that we transition devel list, and eventually most of our mailing lists, to Fedora Discussion (our Discourse-powered forum).
You are 19 days late for April Fools!
Discourse is an absolute pain in the neck because it not only requires JavaScript to be able to participate (the fallback version is read-only), but that JavaScript requires the very latest Chromium or Firefox to work at all. Maintained LTS branches such as QtWebEngine 5.15 LTS are NOT supported and get only the read-only no-JS fallback view: https://discuss.kde.org/t/discuss-kde-org-cannot-be-accessed-by-konqueror/54...
Discourse also does NOT cover the use cases of mailing lists. E-mail notification is just that, notification. Everything still happens on the JavaScript site. And Discourse does NOT support NNTP access, which is the most practical way to interact with a high-volume mailing list such as this one. I am writing this post through the Gmane NNTP gateway using KNode, and I am also using that to read the threads.
As a result, I consider your proposal absolutely unacceptable.
Kevin Kofler
On Thu, Apr 20, 2023 at 11:21 PM Matthew Miller mattdm@fedoraproject.org wrote:
I propose that we transition devel list, and eventually most of our mailing lists, to Fedora Discussion (our Discourse-powered forum).
I've spent more than a decade perfecting my email filters and I have a setup that works for me very well. I dislike certain aspects of mailing lists (cross-posting, top-posting, reply-to, etc, which just can't work well when everyone has to be vigilant all the time to do things right), but I *like* my existing setup and processes. But that's me, us, the old timers.
Having said that, my impression is that mailing lists are an already lost battle. If you don't have an influx of new contributors, your project is going to die eventually. Mailing lists are a big hurdle for newcomers. Young people are not used to it (who still uses mailing lists, in read-write mode, except for OSS communities?), the lists are difficult to set up, the user interfaces are bad, there are many peculiarities to be aware of (top-posting, etc). I don't know if moving away from mailing lists will make our contributor base grow, but I'm quite certain that staying with mailing lists will make our contributor base **not** grow. And our project will slowly decline over time.
Imagine that you want to contribute to a project and you discover they're still using svn, or cvs. There are still some. I personally wouldn't be bothered, I'd just invest time elsewhere. I value my time. I imagine this feeling might be similar to what younger people feel when they're asked to use a mailing list. It would require too much investment from them, and so they'll go somewhere else with a more comfortable barrier to entry.
With some things, there's a simple transition path, because new things are both technically superior and more comfortable at the same time, and still extremely similar. Svn -> git. Plain git -> git with pull requests. Irc -> Matrix. But we don't really have the same equivalent with mailing lists. I guess mailman3+hyperkitty was supposed to be that replacement, but as we can see, it didn't work out. A discussion forum is very different, but it's a workflow that people are familiar with and has much easier onboarding. Mailing lists are a lost cause. They are being abandoned, nobody migrates *to* mailing lists. If we want to keep new blood flowing in, we can't be afraid to change stuff. We have to adapt, instead of requiring everyone else to adapt to us.
I already have some experience with Discourse and so far it seemed OK to me. But I haven't used it as frequently and in such a volume as mailing lists. I also still haven't built my workflow alternative to my email workflow in there. But I'd be happy to experiment with it. But in order to get real-world experience, it would be great if this was a shared experience, where a mass of users really moves and starts using it as a primary discussion platform. Then we can collectively figure out the benefits and drawbacks and any workarounds for those drawbacks.
Cheers, Kamil
Kamil Paral wrote:
I've spent more than a decade perfecting my email filters and I have a setup that works for me very well. I dislike certain aspects of mailing lists (cross-posting, top-posting, reply-to, etc, which just can't work well when everyone has to be vigilant all the time to do things right), but I *like* my existing setup and processes. But that's me, us, the old timers.
But that is exactly why it is an absurd idea to move away from mailing lists. Fedora will *lose* all the existing contributors like you or me.
Having said that, my impression is that mailing lists are an already lost battle. If you don't have an influx of new contributors, your project is going to die eventually. Mailing lists are a big hurdle for newcomers. Young people are not used to it (who still uses mailing lists, in read-write mode, except for OSS communities?), the lists are difficult to set up, the user interfaces are bad, there are many peculiarities to be aware of (top-posting, etc). I don't know if moving away from mailing lists will make our contributor base grow, but I'm quite certain that staying with mailing lists will make our contributor base **not** grow. And our project will slowly decline over time.
I strongly doubt that the mailing list is the main barrier to entry to Fedora (as opposed to, e.g., packaging guidelines, etc.).
If the issue is that the mailing list is flooding your inbox and you are unable to filter it, the remedy is simple: Disable mail delivery and use Gmane NNTP or HyperKitty to read and post to the mailing list. That choice of preferred technology will go away if we move to a locked-in web platform such as Discourse. (I know it is FOSS and the data can be exported somehow. It is still a lock-in for the end user.)
Imagine that you want to contribute to a project and you discover they're still using svn, or cvs. There are still some. I personally wouldn't be bothered, I'd just invest time elsewhere. I value my time.
I do not see the problem. I just need to bring up another UI (Kdesvn or Cervisia instead of Git-Cola), so what? All I care is that I can update from upstream and commit/push my changes.
Heck, I still use SVN for *my* personal projects.
And not everything is better with git. SVN basically guarantees a linear history whereas with git, I have to follow a specific workflow for that (always pull with rebase, never with the default merge strategy, and for work branches, always rebase and force-push them rather than merging master/main into them, then when done fast-forward them to the master/main), and when working with other people, I usually cannot get them to use it, so the history becomes a complicated DAG with a mess of merge commits, grrr!
I already have some experience with Discourse and so far it seemed OK to me.
I have some experience with Discourse as well and it is just a pain: * A silly reliance on JavaScript and AJAX for everything instead of server- side code. In particular, this means it takes several seconds to render a page on mobile devices such as the PinePhone. * A constant requirement of the latest&greatest bleeding edge browser, no support for QtWebEngine LTS branches. * An absurd assumption that everyone is new to the Internet, leading to lots of ridiculous gamified spam "achievements" for basic things such as replying to a thread, with which you get bothered on every single Discourse forum you (have to) sign up to. etc.
I do not understand why everybody is moving to this annoying piece of software and throwing away not only mailing lists, but also web forums using much better software (i.e., pretty much any other forum software), for it.
But I haven't used it as frequently and in such a volume as mailing lists. I also still haven't built my workflow alternative to my email workflow in there. But I'd be happy to experiment with it. But in order to get real-world experience, it would be great if this was a shared experience, where a mass of users really moves and starts using it as a primary discussion platform. Then we can collectively figure out the benefits and drawbacks and any workarounds for those drawbacks.
The drawbacks are well known (see above) and the platform is basically unusable. There is no need to experiment with it to find that out. And "get[ting] real-world experience" by forcing everyone to use it "as a primary discussion platform" is just unacceptable.
Kevin Kofler
On Tue, Apr 25, 2023 at 4:21 AM Kevin Kofler via devel < devel@lists.fedoraproject.org> wrote:
- An absurd assumption that everyone is new to the Internet, leading to
lots of ridiculous gamified spam "achievements" for basic things such as replying to a thread, with which you get bothered on every single Discourse forum you (have to) sign up to.
This is a part I agree with. I've been very annoyed by the gamified achievements, and I haven't found a way to completely shut them off. Fortunately they stopped appearing after some time, perhaps I earned all of them already. Or perhaps they were enabled on Ask, but they're not enabled on Discussions?
Matt, is this something configurable?
On Tue, Apr 25, 2023 at 02:00:02PM +0200, Kamil Paral wrote:
This is a part I agree with. I've been very annoyed by the gamified achievements, and I haven't found a way to completely shut them off. Fortunately they stopped appearing after some time, perhaps I earned all of them already. Or perhaps they were enabled on Ask, but they're not enabled on Discussions?
Matt, is this something configurable?
Yes, I turned them all off on Discussion. (Except the one for having 10+ accepted answers in Ask.) My intention (in very hacky prototype currently) is to replace those all with Fedora Badges.
These will only appear for people who have enabled that (although we may want it to default to on eventually, once Badges is in better shape). And while I hope to have some for Discourse participation, those will be part of the overall Fedora Project ones. (Including replacing the 10+ answers one.)
On Mon, Apr 24, 2023 at 10:21 PM Kevin Kofler via devel devel@lists.fedoraproject.org wrote:
Kamil Paral wrote:
I've spent more than a decade perfecting my email filters and I have a setup that works for me very well. I dislike certain aspects of mailing lists (cross-posting, top-posting, reply-to, etc, which just can't work well when everyone has to be vigilant all the time to do things right), but I *like* my existing setup and processes. But that's me, us, the old timers.
But that is exactly why it is an absurd idea to move away from mailing lists. Fedora will *lose* all the existing contributors like you or me.
I want to make sure I understand that statement. You're saying you will actively walk away from Fedora because you would have to change the manner in which you discuss things?
That's certainly a choice one could make. Personally, I would not prioritize keeping my bespoke email setup intact over working with a community on a project I care about. If there's even a remote possibility moving to Discourse will attract more contributors to Fedora then I'd happily learn how to deal with it. Change is difficult, but in the end it's something we're all capable of.
josh
On Tue, Apr 25, 2023 at 02:59:22PM -0400, Josh Boyer wrote:
That's certainly a choice one could make. Personally, I would not prioritize keeping my bespoke email setup intact over working with a community on a project I care about.
My "besoke email setup" benefits *me* and my productivity/workflow far more than the ability to interact with Fedora. I might add that "email" as a whole remains a _lot_ more important than Fedora too.
For nearly all of us, Fedora is means to an end, not the end in of itself. Please keep that in mind that most of us are _not_ paid or otherwise compensated for our participation here, and that there are very real costs (if "only" time and attention) associated with this participating. By increasing those costs (real and perceived) many will reasonably determine the benefits are no longer worth the higher costs.
If there's even a remote possibility moving to Discourse will attract more contributors to Fedora then I'd happily learn how to deal with it. Change is difficult, but in the end it's something we're all capable of.
What's good for "Fedora" isn't necessarily good for many (or even most) of the individuals that already participate.
Again, this is a matter of deciding, at the project leadership level, what level of effort should go into keeping more of the current contributors versus the possibility of attracting new ones. As things stand, the leadership is tryting hard to understand the objections to/problems with this proposal and mitigate them "well enough"
Honestly, if a "how to configure discourse to mimic the MUA-managed mailing list experience (ie not having to log into a web site after the initial configuration)" document is produced, that's probaby sufficient to overcome most of these objections, because then the setup cost is one-off, and the ongoing "interact with Fedora-devel" cost won't be any greater than it already is. (It's not the setup cost that's a problem, it's the per-transaction/interaction cost, which is currently _higher_)
- Solomon
On 4/25/23 14:33, Solomon Peachy via devel wrote:
Honestly, if a "how to configure discourse to mimic the MUA-managed mailing list experience (ie not having to log into a web site after the initial configuration)" document is produced, that's probaby sufficient to overcome most of these objections, because then the setup cost is one-off, and the ongoing "interact with Fedora-devel" cost won't be any greater than it already is.
Only if there's a companion document on how to interact with Discourse over NNTP. :-(
On 4/25/23 14:33, Solomon Peachy via devel wrote:
Honestly, if a "how to configure discourse to mimic the MUA-managed mailing list experience (ie not having to log into a web site after the initial configuration)" document is produced, that's probaby sufficient to overcome most of these objections, because then the setup cost is one-off, and the ongoing "interact with Fedora-devel" cost won't be any greater than it already is.
As far as I can tell, it is impossible to configure Discourse to work like a mailing list. It is not designed for that. It is a web AJAX forum, with the possibility to get e-mail notifications, but the e-mail interaction is limited and cannot replace the AJAX UI.
And in particular, NNTP is not supported at all, and the way the e-mail notifications work does not lend itself to NNTP gatewaying over Gmane.
Ian Pilcher wrote:
Only if there's a companion document on how to interact with Discourse over NNTP. :-(
Exactly that.
Kevin Kofler
On Wed, Apr 26, 2023 at 04:34:22AM +0200, Kevin Kofler via devel wrote:
And in particular, NNTP is not supported at all, and the way the e-mail notifications work does not lend itself to NNTP gatewaying over Gmane.
Well, email (and Mailman3 and Hyperkitty and Postorious) do not support NNTP. Gmane is a software that specifically exists to bridge email to NNTP.
I actually love the idea of a Discourse-to-NNTP bridge.
Matthew Miller wrote:
I actually love the idea of a Discourse-to-NNTP bridge.
But is it ever going to happen? What is sure is that Discourse's e-mail notification system cannot be used for this, or at least it is not suitable for the existing e-mail to NNTP bridges. So somebody would need to write a dedicated bridge.
Kevin Kofler
On Tue, Apr 25, 2023 at 03:33:28PM -0400, Solomon Peachy via devel wrote:
Honestly, if a "how to configure discourse to mimic the MUA-managed mailing list experience (ie not having to log into a web site after the initial configuration)" document is produced, that's probaby sufficient to overcome most of these objections, because then the setup cost is one-off, and the ongoing "interact with Fedora-devel" cost won't be any greater than it already is. (It's not the setup cost that's a problem, it's the per-transaction/interaction cost, which is currently _higher_)
I've written this:
https://discussion.fedoraproject.org/t/guide-to-interacting-with-this-site-b...
and I'd love feedback on it.
Keep in mind:
* It never will be the same as mailman. There are different constraints.
* Discourse developers and designers see the web interaction as primary, and that isn't going to change.
* Dealing with email has a lot of corner cases, so there will be rough edges.
Matthew Miller wrote:
I've written this:
https://discussion.fedoraproject.org/t/guide-to-interacting-with-this-site-b...
and I'd love feedback on it.
Asking feedback from users who are not using Discourse… on Discourse (!) is absurd and the best way to not get any answers. (It is like asking "Everyone who does not speak English, please raise your hand!" in English.)
Note that we cannot use the reply by e-mail feature to answer in the thread because it is an already existing thread and so we will not get a notification to reply to if we turn on e-mail notifications now.
Keep in mind:
- It never will be the same as mailman. There are different constraints.
And that is exactly why Discourse will by design never be a replacement for a mailing list.
- Discourse developers and designers see the web interaction as primary, and that isn't going to change.
And that too.
- Dealing with email has a lot of corner cases, so there will be rough edges.
And that is a logical consequence of having the web forum as the primary interface and mail as an afterthought rather than the opposite.
Some features such as editing posts just by design cannot work on a mailing list.
Kevin Kofler
On Thu, Apr 27, 2023 at 12:49:02AM +0200, Kevin Kofler via devel wrote:
Matthew Miller wrote:
I've written this: https://discussion.fedoraproject.org/t/guide-to-interacting-with-this-site-b... and I'd love feedback on it.
Asking feedback from users who are not using Discourse… on Discourse (!) is absurd and the best way to not get any answers. (It is like asking "Everyone who does not speak English, please raise your hand!" in English.)
I'm sorry, I guess my phrasing was not clear. I would like feedback _about_ it. That feedback can be here or directly emailed to me.
On Wed, Apr 26, 2023 at 10:28:41AM -0400, Matthew Miller wrote:
On Tue, Apr 25, 2023 at 03:33:28PM -0400, Solomon Peachy via devel wrote:
Honestly, if a "how to configure discourse to mimic the MUA-managed mailing list experience (ie not having to log into a web site after the initial configuration)" document is produced, that's probaby sufficient to overcome most of these objections, because then the setup cost is one-off, and the ongoing "interact with Fedora-devel" cost won't be any greater than it already is. (It's not the setup cost that's a problem, it's the per-transaction/interaction cost, which is currently _higher_)
I've written this:
https://discussion.fedoraproject.org/t/guide-to-interacting-with-this-site-b...
and I'd love feedback on it.
I turned on the full email notications for "Project discussion" to get a representative sample. After two days, I would say that those notifications are not perfect, e.g. the quoting is off for no good reason [1], but in general OK. The text is readable and would work okayish to just follow the discussions. In particular, replies to replies seems to be threaded correctly, so I see a little tree in mutt. (In some previous comments on this thread, people who passively follow fedora-devel were mentioned. I think this answers this particular concern.)
Personally, I expect this would be similar to how I interact with github: I get notifications by email for various activities, and either delete them or do short replies using email, but for "bigger" interactions I click on one of the links to go the web forum. The email workflow could be made better, github suffers from the same gratuitous bad plain text formatting, but practically there are limits: a multi-page diff cannot be displayed in the plain-text email anyway, so some interaction through the browser would be needed anyway.
Actually, I think the metadata that discourse sends is already better than what github does. And the formatting is also better. If [1] is fixed, I think it would rate as "good". The threading is definitely better in discourse than in github, because the latter always does a flat list of replies.
[1] https://meta.discourse.org/t/format-quotes-in-plain-text-e-mail-correctly/26...
Zbyszek
Josh Boyer wrote:
I want to make sure I understand that statement. You're saying you will actively walk away from Fedora because you would have to change the manner in which you discuss things?
I am saying I and many other existing contributors may or may not walk away from Fedora entirely, and even if not, may reduce our interaction with Fedora, due to being forced to use a discussion tool that does not support our workflows.
Heck, it does not even work in my browser (Falkon) without ugly workarounds. (https://discuss.kde.org/t/discuss-kde-org-cannot-be-accessed-by-konqueror/54...)
That's certainly a choice one could make. Personally, I would not prioritize keeping my bespoke email setup intact over working with a community on a project I care about. If there's even a remote possibility moving to Discourse will attract more contributors to Fedora then I'd happily learn how to deal with it. Change is difficult, but in the end it's something we're all capable of.
I do not understand this double standard: You and several others are expecting new contributors to walk away from Fedora due to the manner in which we discuss things, but in the same time act surprised when warned that existing contributors are likely to do exactly that. In fact, we have much more reason to do so because the new workflow is a regression, and forcing it on us over our objections actively tells us we are no longer welcome.
Kevin Kofler
On Tue, Apr 25, 2023 at 10:42 PM Kevin Kofler via devel devel@lists.fedoraproject.org wrote:
Josh Boyer wrote:
I want to make sure I understand that statement. You're saying you will actively walk away from Fedora because you would have to change the manner in which you discuss things?
I am saying I and many other existing contributors may or may not walk away from Fedora entirely, and even if not, may reduce our interaction with Fedora, due to being forced to use a discussion tool that does not support our workflows.
Ah. May or may not. That gives me hope at least.
Heck, it does not even work in my browser (Falkon) without ugly workarounds. (https://discuss.kde.org/t/discuss-kde-org-cannot-be-accessed-by-konqueror/54...)
That's certainly a choice one could make. Personally, I would not prioritize keeping my bespoke email setup intact over working with a community on a project I care about. If there's even a remote possibility moving to Discourse will attract more contributors to Fedora then I'd happily learn how to deal with it. Change is difficult, but in the end it's something we're all capable of.
I do not understand this double standard: You and several others are expecting new contributors to walk away from Fedora due to the manner in which we discuss things, but in the same time act surprised when warned that existing contributors are likely to do exactly that. In fact, we have much more reason to do so because the new workflow is a regression, and forcing it on us over our objections actively tells us we are no longer welcome.
I don't think it's a double standard. I think people that already know how Fedora development and contribution works are inherently in a better position to adapt to something new, whereas net-new contributors are unlikely to even start based on an email driven practice. It's a barrier to entry problem. If we, as the experienced and capable contributor base, can adapt to something and lower the barrier to entry then it benefits us all.
To be clear, I do not like forums. Discourse is perhaps the best forum platform I've used, but that doesn't mean I like it. What I dislike more than forum software is knowing we're leaving people that could do great things and spread more awareness of Fedora simply because we've ossified on a communication platform that is hard to discover and consume.
josh
Josh Boyer wrote:
Ah. May or may not. That gives me hope at least.
Well, considering that we have hundreds of existing contributors, who all may or may not be willing to adapt to a platform that is clearly not designed for them (Discourse is very strongly newbie-centric, see the "achievements" and all the other hand-holding), I think it is safe to assume that several important contributors WILL leave, tone down their participation, and/or miss some important communication (leading to breakage in the distribution, e.g., broken dependencies making it into Rawhide) if Fedora makes the switch.
I don't think it's a double standard. I think people that already know how Fedora development and contribution works are inherently in a better position to adapt to something new, whereas net-new contributors are unlikely to even start based on an email driven practice.
And as I already answered, I think this is completely backwards. If you want to newly join a project, you learn their way to do things and adapt to it. If, on the other hand, you are already involved in a project and have workflows that work for you, any forced change to something perceived as a regression will annoy you and make you want to leave.
It's a barrier to entry problem. If we, as the experienced and capable contributor base, can adapt to something and lower the barrier to entry then it benefits us all.
Again, I do not see the communication platform as the main barrier to entry for Fedora at all. Where is the evidence for that claim?
As I see it, the main roadblocks for new packagers are: * accepting the FPCA, * getting sponsored, * learning the Packaging Guidelines, and * getting their package(s) through review, and that last point can be a roadblock even for existing packagers (because we do not trust even experienced provenpackagers and/or packager sponsors to review their own packages).
Those points are all there for a reason (the FPCA for legal coverage, the sponsorship process so new contributors are mentored and their trustfulness verified, the Packaging Guidelines to ensure a certain package quality for our end users, and the review process to ensure that the Packaging Guidelines are actually followed and as another check that nothing malicious sneaks in), but they are the barrier to entry, not the communication platform.
To be clear, I do not like forums. Discourse is perhaps the best forum platform I've used, but that doesn't mean I like it.
I find Discourse to actually be one of the worst forum platforms I have used, if not the worst. All the *BB ones out there are much better due to using less (to no) AJAX and less annoying hand-holding.
What I dislike more than forum software is knowing we're leaving people that could do great things and spread more awareness of Fedora simply because we've ossified on a communication platform that is hard to discover and consume.
Again, this assumes that this is what is holding back new contributors, for which I have seen no evidence whatsoever.
Kevin Kofler
On Wed, Apr 26, 2023 at 02:42:56PM +0200, Kevin Kofler via devel wrote:
And as I already answered, I think this is completely backwards. If you want to newly join a project, you learn their way to do things and adapt to it. If, on the other hand, you are already involved in a project and have workflows that work for you, any forced change to something perceived as a regression will annoy you and make you want to leave.
In nearly all professions, "mastery of the tools of the trade" is considered to be an important milestone, necessary if you're wanting to excel in that endeavor. Unfortunately, for some reason, in this profession, "tool mastery" is actively looked down upon. Nevermind that this mastery (which includes continuing to adapt, refine, and improve those tools) is the basis of the "10x developer" that so many claim to want.
As I see it, the main roadblocks for new packagers are:
I think one of the legitimate points Matthew (and others) are making is that "contributor != packager"
- accepting the FPCA,
- getting sponsored,
FWIW I never made it past this point. But at the same time I've not really needed to.
- learning the Packaging Guidelines, and
- getting their package(s) through review,
But this is a _very_ high hurdle, as there's a _ton_ of policy and fedora-specific process that has nothing to do with email or forums or any "common" communication platform that would-be contributors would ever be expected to be familiar with.
But as I already mentioned, this only applies to packagers; there are many other avenues for contributors. That said, are those other avenues relevant for this particular mailing list?
Guidelines are actually followed and as another check that nothing malicious sneaks in), but they are the barrier to entry, not the communication platform.
I completely agree, at least on the "packagers" side. I don't know what barriers non-packaging contributors face, but I presume each sub-category has its own barriers to entry.
- Solomon
On Wed, 26 Apr 2023 at 08:44, Kevin Kofler via devel < devel@lists.fedoraproject.org> wrote:
Josh Boyer wrote:
Ah. May or may not. That gives me hope at least.
Well, considering that we have hundreds of existing contributors, who all may or may not be willing to adapt to a platform that is clearly not designed for them (Discourse is very strongly newbie-centric, see the "achievements" and all the other hand-holding), I think it is safe to assume that several important contributors WILL leave, tone down their participation, and/or miss some important communication (leading to breakage in the distribution, e.g., broken dependencies making it into Rawhide) if Fedora makes the switch.
I may be in that list of toning down, but that is OK. Look it's really time for new people to come in and break things. It is the only way they really learn if something is something that should be really avoided or was a taboo we had from the 1990's which we can't see as cargo culting anymore.
Maybe a bunch of packages will be dropped and Fedora will become 'useless' to some of us older contributors. This isn't the first time that has happened with the distribution (we saw large drop-offs after we stopped Xen and when we changed desktops to GNOME3.) and if the distribution is to last as an institution, it won't be the last. We who aren't happy with it can either make do with something else, adapt, or finally retire to grow potatoes like all the programmers I knew from the 1980s who had gotten tired of all the changes over the years.
Normally this is where I would clip the rest of the message, but I want to say something about a later section, so please scroll.
It's a barrier to entry problem. If we, as the experienced and capable contributor base, can adapt to something and lower the barrier to entry then it benefits us all.
Again, I do not see the communication platform as the main barrier to entry for Fedora at all. Where is the evidence for that claim?
As I see it, the main roadblocks for new packagers are:
- accepting the FPCA,
- getting sponsored,
- learning the Packaging Guidelines, and
- getting their package(s) through review,
and that last point can be a roadblock even for existing packagers (because we do not trust even experienced provenpackagers and/or packager sponsors to review their own packages).
Those points are all there for a reason (the FPCA for legal coverage, the sponsorship process so new contributors are mentored and their trustfulness verified, the Packaging Guidelines to ensure a certain package quality for our end users, and the review process to ensure that the Packaging Guidelines are actually followed and as another check that nothing malicious sneaks in), but they are the barrier to entry, not the communication platform.
I want to say that I agree with Kevin Kofler on this. We have a lot of other barriers for entry which I have found to be higher on getting new contributors into the distribution. Unless we have some field in the discussion site with interested and energized people who can help mentor future packagers.. we aren't addressing the real problem. That said, this discussion hasn't been about how to fix that problem either here OR the forum.
Stephen Smoogen wrote:
I may be in that list of toning down, but that is OK. Look it's really time for new people to come in and break things. It is the only way they really learn if something is something that should be really avoided or was a taboo we had from the 1990's which we can't see as cargo culting anymore.
Maybe a bunch of packages will be dropped and Fedora will become 'useless' to some of us older contributors. This isn't the first time that has happened with the distribution (we saw large drop-offs after we stopped Xen and when we changed desktops to GNOME3.) and if the distribution is to last as an institution, it won't be the last. We who aren't happy with it can either make do with something else, adapt, or finally retire to grow potatoes like all the programmers I knew from the 1980s who had gotten tired of all the changes over the years.
But what if the major influx of new contributors that you proponents of this proposal are hoping for never arrives? (Something I think is quite likely to happen, considering that the main barrier to entry is NOT the mailing list.) Then you will have driven away existing key contributors without anyone to replace them, and Fedora will be dead.
Also keep in mind that experienced contributors are the only ones able to work on certain complex tasks and also to mentor new contributors so that they will eventually become experienced. Chase them away and all the experience will be lost, no matter how many new contributors you attract.
The first priority of a project MUST ALWAYS be to keep the existing contributors. Attracting new ones can only come second.
That said, looking at how the "feedback" to this "proposal" is being handled (yet again), I guess that all that is really going to change in practice is that instead of completely ignoring or trying to explain away all mailing list feedback, you folks will be completely ignoring or trying to explain away all Discourse feedback. How proposals are supposed to work is that somebody suggests something, then feedback is requested, then it is discussed, and only then a decision is made. But how it is actually working is that a small group of people decides something in a closed-door meeting, then calls it a "proposal", asks for "feedback" on the mailing list to give an illusion of transparency and democracy, but is not actually willing to act on the feedback (because the decision has long been made elsewhere), instead only trying to explain why it "does not matter".
The sad thing is that I have even seen several replies trying to explain away their OWN objections to the proposal, arguing that it is normal that they as experienced contributors are not the target user base of the discussion platform. But guess what, it is not (normal). This is a developer mailing list, experienced contributors are THE target.
Kevin Kofler
On Wed, 26 Apr 2023 at 23:58, Kevin Kofler via devel < devel@lists.fedoraproject.org> wrote:
Stephen Smoogen wrote:
I may be in that list of toning down, but that is OK. Look it's really time for new people to come in and break things. It is the only way they really learn if something is something that should be really avoided or was a taboo we had from the 1990's which we can't see as cargo culting anymore.
Maybe a bunch of packages will be dropped and Fedora will become
'useless'
to some of us older contributors. This isn't the first time that has happened with the distribution (we saw large drop-offs after we stopped Xen and when we changed desktops to GNOME3.) and if the distribution is
to
last as an institution, it won't be the last. We who aren't happy with it can either make do with something else, adapt, or finally retire to grow potatoes like all the programmers I knew from the 1980s who had gotten tired of all the changes over the years.
But what if the major influx of new contributors that you proponents of this
s/you/the/;
I am NOT a proponent of this proposal. I don't want to go to Discourse. Web interfaces like that cause me cognitive pain and grumpiness to use longer than a few minutes. As such I know my involvement with Fedora will go down further.
If it comes across that I am for this change, it is because I am tired and frustrated. The mailman system has been running on inertia since at least February 2018, when the last software updates to the mailman software were done. Over the last 5 years, the system has mostly run, but in the last year has increasingly had longer and longer outages. My tiredness comes from spending most of my Thanksgiving and Winter breaks trying to find reasons and then doing whatever cave-man hacks I could to fix it without breaking mail altogether. My frustration and anger comes from the fact that I spent most of the last 5 years assuming that it was somebody else's problem and they would take care of it so I could focus on keeping other things running.
On Thu, Apr 27, 2023 at 07:01:15AM -0400, Stephen Smoogen wrote:
breaking mail altogether. My frustration and anger comes from the fact that I spent most of the last 5 years assuming that it was somebody else's problem and they would take care of it so I could focus on keeping other things running.
This is a very important point -- Is RH/Fedora prepared to properly handle the maintainence, administrative, moderation, etc burden of scaling up the Discourse instance?
Or will all of Fedora's customizations make it into another special snowflake instance that results in very painful upgrade paths, leaving it to become yet another service left to coast along under its own inertia until this cycle repeats itself again?
I mean, it's all fine to say "but Discourse is actively developed" -- if you never actually upgrade/update it to match upstream, it's no different than the situation we have with our mailman3 today, where we're literally years behind the curve.
- Solomon
On Thu, 27 Apr 2023 07:01:15 -0400 Stephen Smoogen ssmoogen@redhat.com wrote:
I am NOT a proponent of this proposal. I don't want to go to Discourse. Web interfaces like that cause me cognitive pain and grumpiness to use longer than a few minutes. As such I know my involvement with Fedora will go down further.
If it comes across that I am for this change, it is because I am tired and frustrated. The mailman system has been running on inertia since at least February 2018, when the last software updates to the mailman software were done. Over the last 5 years, the system has mostly run, but in the last year has increasingly had longer and longer outages. My tiredness comes from spending most of my Thanksgiving and Winter breaks trying to find reasons and then doing whatever cave-man hacks I could to fix it without breaking mail altogether. My frustration and anger comes from the fact that I spent most of the last 5 years assuming that it was somebody else's problem and they would take care of it so I could focus on keeping other things running.
I know almost nothing about mail list infrastructure.
Are there other open source linux distributions using the latest mailman? Could their process be copied and put in as a drop in replacement for fedora with a little tweaking? How do they deal with the spam problem? I'm not asking you to do it, but you appear to be a domain expert, so you can probably answer these questions off the top of your head.
Some more questions. What language is mailman written in? What are the major incompatibilities of the new version with the older version? Are there more modern alternatives that are easier to set up and maintain?
Crazy ideas.
Would it cost less resources to set up a private usenet server for messages than to continue maintaining the mailman application? Is it even possible? Could it be outsourced to one of the big usenet providers? I doubt it would be even a dent in their capacity since they maintain 10 years of binary usenet posts. Could such a usenet server be kept in sync with Discourse, and allow posts to the server to be propagated to Discourse? I'm thinking of some kind of automated web application that takes each post on the usenet server and logs in to Discourse and posts it to the appropriate place. A user would configure it with their credentials once, and forget it. Sort of the way youtube-downloader works, except opposite. In the other direction, there would be a web scraper that regularly scrapes posts on fedora discourse and reposts them to the usenet server. Again, the user enters their credentials once, and done.
Ignore this if it seems too woo woo or irrelevant or resource intense.
Forgive me for answering 3 posts in one here...
On Thu, Apr 27, 2023 at 05:57:43AM +0200, Kevin Kofler via devel wrote:
But what if the major influx of new contributors that you proponents of this proposal are hoping for never arrives? (Something I think is quite likely to happen, considering that the main barrier to entry is NOT the mailing list.) Then you will have driven away existing key contributors without anyone to replace them, and Fedora will be dead.
I don't see a 'we are all in on discourse, please turn off the mailing lists' until there's a critical mass of discussion over on discourse. Perhaps thats just me, I can't speak for the council, but if things don't go well at some point, the lists will still be here.
Also keep in mind that experienced contributors are the only ones able to work on certain complex tasks and also to mentor new contributors so that they will eventually become experienced. Chase them away and all the experience will be lost, no matter how many new contributors you attract.
The first priority of a project MUST ALWAYS be to keep the existing contributors. Attracting new ones can only come second.
I think you are conflating contributors with "packagers". Granted packagers are super important, but there's a ton of other contributors out there. People writing tests, making documentation, etc. Also, improving things here for everyone who wants to talk about discussion doesn't mean we shouldn't try and improve other parts of being a packager, and in fact there's things ongoing to do that. ...snip...
On Thu, Apr 27, 2023 at 10:07:42AM -0400, Solomon Peachy via devel wrote:
On Thu, Apr 27, 2023 at 07:01:15AM -0400, Stephen Smoogen wrote:
breaking mail altogether. My frustration and anger comes from the fact that I spent most of the last 5 years assuming that it was somebody else's problem and they would take care of it so I could focus on keeping other things running.
This is a very important point -- Is RH/Fedora prepared to properly handle the maintainence, administrative, moderation, etc burden of scaling up the Discourse instance?
Our discourse instance is hosted for us by discourse. We shouldn't have to do maint on it, but we will have to do moderation, etc.
Or will all of Fedora's customizations make it into another special snowflake instance that results in very painful upgrade paths, leaving it to become yet another service left to coast along under its own inertia until this cycle repeats itself again?
I mean, it's all fine to say "but Discourse is actively developed" -- if you never actually upgrade/update it to match upstream, it's no different than the situation we have with our mailman3 today, where we're literally years behind the curve.
They are managing it for us. We don't have to do anything except pay them. :)
Also, we do have full backups of the database and assets. In the unlikely event that we needed to pull out our data, we could (all be it with some pain trying to export it into another system).
kevin
On Thu, Apr 27, 2023 at 10:47:15AM -0700, Kevin Fenzi wrote:
Our discourse instance is hosted for us by discourse. We shouldn't have to do maint on it, but we will have to do moderation, etc.
So... if maintaining discourse is too much overhead but it's okay to pay someone else to handle, why can't that be done for our mailing list infrastructure too? Even if that service offering is proprietary? (hosted enterprise gitlab, anyone?)
And, for that matter, what of the other services currently owned/hosted/maintained under the RH/Fedora roof? (For example, we're going to be having this conversation again in the not-so-distant future once RH finishes switching over to Jira and stops funding our Bugzilla instance's upkeep..)
- Solomon
On Fri, Apr 28, 2023 at 08:15:28AM -0400, Solomon Peachy via devel wrote:
On Thu, Apr 27, 2023 at 10:47:15AM -0700, Kevin Fenzi wrote:
Our discourse instance is hosted for us by discourse. We shouldn't have to do maint on it, but we will have to do moderation, etc.
So... if maintaining discourse is too much overhead but it's okay to pay someone else to handle, why can't that be done for our mailing list infrastructure too? Even if that service offering is proprietary? (hosted enterprise gitlab, anyone?)
We could, but as I have mentioned a number of times... it's not about _our mailing lists_ it's about mailing lists in general.
Also, as Matthew mentioned somewhere in this megathread, it's actually a difficult process to add a new vendor at Red Hat.
And, for that matter, what of the other services currently owned/hosted/maintained under the RH/Fedora roof? (For example, we're going to be having this conversation again in the not-so-distant future once RH finishes switching over to Jira and stops funding our Bugzilla instance's upkeep..)
Yep, someday I think we will. There is currently no plan I know about to retire bugzilla, but there could well be someday.
kevin
Kevin Fenzi wrote:
I think you are conflating contributors with "packagers".
We are talking about moving things off the devel list, which is mainly a packaging mailing list.
Our discourse instance is hosted for us by discourse. We shouldn't have to do maint on it, but we will have to do moderation, etc.
They are managing it for us. We don't have to do anything except pay them. :)
So it is effectively not Free Software for us, given that we use it as SaaS.
Kevin Kofler
[long discussion elided]
It has to be said I'm very much agreeing with Kevin Kofler on this.
Rich.
On Wed, Apr 26, 2023 at 8:43 AM Kevin Kofler via devel devel@lists.fedoraproject.org wrote:
Josh Boyer wrote:
Ah. May or may not. That gives me hope at least.
Well, considering that we have hundreds of existing contributors, who all may or may not be willing to adapt to a platform that is clearly not designed for them (Discourse is very strongly newbie-centric, see the "achievements" and all the other hand-holding), I think it is safe to assume that several important contributors WILL leave, tone down their participation, and/or miss some important communication (leading to breakage in the distribution, e.g., broken dependencies making it into Rawhide) if Fedora makes the switch.
I don't think it's a double standard. I think people that already know how Fedora development and contribution works are inherently in a better position to adapt to something new, whereas net-new contributors are unlikely to even start based on an email driven practice.
And as I already answered, I think this is completely backwards. If you want to newly join a project, you learn their way to do things and adapt to it. If, on the other hand, you are already involved in a project and have workflows that work for you, any forced change to something perceived as a regression will annoy you and make you want to leave.
I think you and I fundamentally disagree on this and that's OK. I'm not trying to convince you. I was trying to understand your stance on why you'd potentially leave and express the reasons why I think it's worth it.
We'll have to end our conversation here, because we've both stated clearly where we stand and I respect that you and I are simply not aligned. Thank you for your thoughts.
josh
On 4/26/23 08:42, Kevin Kofler via devel wrote:
As I see it, the main roadblocks for new packagers are:
- accepting the FPCA,
- getting sponsored,
- learning the Packaging Guidelines, and
- getting their package(s) through review,
and that last point can be a roadblock even for existing packagers (because we do not trust even experienced provenpackagers and/or packager sponsors to review their own packages).
I agree with Kevin on these being much larger roadblocks than e-mail vs. forums.
I also contribute to another distribution, and they accept pull requests on GitHub with the standard git Signed-Off-By tag (i.e., nothing like the FPCA to agree to, no sponsorship needed, really no commitment of any kind needed - just send a PR and wait for someone to review/merge it. It's super easy, barely an inconvenience!) For those who aren't keen on GitHub, the same distro also accepts patches to their mailing list as long as you do the same Signed-Off-By tag. I've found contributing there to be quite easy.
I harbor doubt that changing to a web forum will make a notable difference in technical contributions to this project (the sorts of things that this list is about). I don't carry that doubt about many parts of the Fedora community, so I mean here to focus specifically on what this specific list, devel, is usually about.
I do think we can do a lot to lower that barrier to entry for new comers regarding the things Kevin is talking about, though I suppose the FPCA thing can only be relaxed with lawyer permission. I think it's possible to mirror to GitHub and require FPCA (if we really must have it) - I've seen a few projects on GitHub that had a mechanism to enforce agreeing to a document before they would accept PRs. Loads of people have a GitHub account already and it would be much easier for them than having a FAS account (the other distro I contribute to doesn't require contributors to have an account, though I do).
I think reviewing code is a healthy practice. One thing I've always thought was a little weird is that we only require new packages to be reviewed, but after that the packager can do anything. I guess I've always assumed that this was more about getting a second person to agree to the license/patent implications of the package, and probably not the code itself (since that will change over time without further review). I guess if that's why it kinda makes sense.
On Mon, 24 Apr 2023 19:47:07 +0200 Kamil Paral kparal@redhat.com wrote:
On Thu, Apr 20, 2023 at 11:21 PM Matthew Miller mattdm@fedoraproject.org wrote:
I propose that we transition devel list, and eventually most of our mailing lists, to Fedora Discussion (our Discourse-powered forum).
project is going to die eventually. Mailing lists are a big hurdle for newcomers. Young people are not used to it (who still uses mailing lists, in read-write mode, except for OSS communities?), the lists are difficult to set up, the user interfaces are bad, there are many peculiarities to be aware of (top-posting, etc).
I think an analogy would be that mailing lists are like vim or emacs, and web forums are like nano. To someone who is used to using a power editor, using nano is irritating. Sure they can accomplish the task, but it is like slogging through mud instead of running on a track. If all you've ever known is slogging through mud, of course, that is going to be acceptable, and using a power editor is going to be like slogging through mud (because of the learning curve).
I think that is why there are so many complaints about the switch from existing mail list users. They have a system in place that allows them to run, and really don't want to slog through mud instead.
The change to a forum doesn't really benefit them. The benefits mainly accrue to the maintainers and managers of the forums, not the users. I suppose it could be argued that if the change makes the project more efficient, more viable, they are gaining a benefit, but what assurances are there that the benefit will happen? i.e. what evidence is there that the benefit they are slogging through mud for will actually occur?
I've read all the responses to the announcement, and I don't think I've seen an architectural analysis. What communication requirements does a project like Fedora need? What is the theoretical optimal process for each requirement? Is there an existing optimal tool for each of those optimal processes? Are there sub optimal tools that can be used for multiple requirements? Is the loss of effectiveness / efficiency of using a sub optimal tool worth the reduction in the number of tools?
This is trying to answer the question, "If there was a magic wand we could wave, and perfectly satisfy the communication needs of the fedora project, what would that look like?" And, "How would we get from here to there?"
On Tue, Apr 25, 2023 at 6:29 PM stan via devel devel@lists.fedoraproject.org wrote:
(snip)
I've read all the responses to the announcement, and I don't think I've seen an architectural analysis. What communication requirements does a project like Fedora need? What is the theoretical optimal process for each requirement? Is there an existing optimal tool for each of those optimal processes? Are there sub optimal tools that can be used for multiple requirements? Is the loss of effectiveness / efficiency of using a sub optimal tool worth the reduction in the number of tools?
This is trying to answer the question, "If there was a magic wand we could wave, and perfectly satisfy the communication needs of the fedora project, what would that look like?" And, "How would we get from here to there?"
+1 to this. I am mostly in camp "grumpy old man doesn't want to use new program", but I think we can (and should!) have a discussion about what things happen on the "devel" list. Are there things that are currently handled via the "devel" mailing list that don't necessarily need to be, and for which Discourse might even be a better tool?
Emails sent to the devel list seem to mostly belong to these categories:
- announcement and discussion of Change Proposals - announcement and coordination of breaking changes (like soname bumps / API changes) - discussion of various RFCs / proposals which might eventually become Change proposals - various other announcements (packages {un,}orphaned, packages {un,}retired, etc.) - reports (rawhide compose report, rawhide QA report, SPDX conversion progress, etc.) - introductions of new packagers seeking sponsors
I don't think moving *everything* to Discourse is a good idea, since the mailing list is a better *tool* for handling some of these things - while Discourse might be the better tool for others.
For example (*not serious proposals, just some ideas in an effort to make this discussion more productive and less black-and-white*):
- Change Proposals could be *announced* on the devel list, but discussion could happen in a linked topic on Discourse - announcement and coordination of breaking changes continues to be handled via the mailing list - RFCs and maybe-future-Change-Proposals are discussed on Discourse (maybe with an announcement post on the mailing list) - orphan/unorphan/retire/unretire announcements stay on the devel list - reports stay on the devel list (or are moved to a separate "reports" list that interested parties can subscribe to) - introductions of new packagers happens in the "Introductions" category on Discourse (which I believe already exists?)
Yes, most of these examples would be about moving "discussion-like-things" to discourse while keeping "announcement-like-things" on the mailing list. That should reduce the volume of emails on the list dramatically, make it easier for people to follow, and possibly move actual *discussion* to a tool that was designed to handle *discussion*. It would also make it very easy to subscribe to Discourse topics one is interested in, and ignore others. Just my 2 euro-cents :)
Fabio
On Tue, Apr 25, 2023 at 07:00:20PM +0200, Fabio Valentini wrote:
- Change Proposals could be *announced* on the devel list, but
discussion could happen in a linked topic on Discourse
This is basically my proposal, although I suggest devel-announce rather than devel-list, because otherwise the result is two separate discussions.
Matthew Miller wrote:
On Tue, Apr 25, 2023 at 07:00:20PM +0200, Fabio Valentini wrote:
- Change Proposals could be *announced* on the devel list, but
discussion could happen in a linked topic on Discourse
This is basically my proposal, although I suggest devel-announce rather than devel-list, because otherwise the result is two separate discussions.
devel-announce is forwarded to devel, so this does not prevent the two separate discussions (mailing list vs. Discourse). Unless you stop this forwarding, which would also be a regression because it means we would now have to watch yet another mailing list.
Kevin Kofler
Dne 20. 04. 23 v 23:20 Matthew Miller napsal(a):
I know this is a big change, but, hear me out…
It is a big change. But I am +1
So far I was able to configure notification in Discourse to send me email and I track new in mail client. The Discourse can filter lots of emails that are +1 or -1. And at the same time give you option to provide such feedback in cases you would not submit it as an email.
Miroslav
On 20/04/2023 23:20, Matthew Miller wrote:
It’s time to transform the Fedora devel list into something new
I think such serious questions should be put to a vote. Not a FESCo vote, but a vote for all Fedora contributors (can be combined with the next FESCo elections).
On 4/26/23 16:04, Vitaly Zaitsev via devel wrote:
On 20/04/2023 23:20, Matthew Miller wrote:
It’s time to transform the Fedora devel list into something new
I think such serious questions should be put to a vote. Not a FESCo vote, but a vote for all Fedora contributors (can be combined with the next FESCo elections).
Well it's long overdue that the community in whole cant veto via vote poor decision making by the appointed and elected official of the project including it's leader...
JBG
On Wed Apr 26, 2023 at 18:04 +0200, Vitaly Zaitsev via devel wrote:
On 20/04/2023 23:20, Matthew Miller wrote:
It’s time to transform the Fedora devel list into something new
I think such serious questions should be put to a vote. Not a FESCo vote, but a vote for all Fedora contributors (can be combined with the next FESCo elections).
I think having this as a "ballot referendum" of sorts is a good idea.
-- Maxwell G (@gotmax23) Pronouns: He/They
On Wed, 26 Apr 2023 at 15:28, Maxwell G maxwell@gtmx.me wrote:
On Wed Apr 26, 2023 at 18:04 +0200, Vitaly Zaitsev via devel wrote:
On 20/04/2023 23:20, Matthew Miller wrote:
It’s time to transform the Fedora devel list into something new
I think such serious questions should be put to a vote. Not a FESCo vote, but a vote for all Fedora contributors (can be combined with the next FESCo elections).
I think having this as a "ballot referendum" of sorts is a good idea.
So let us say it is voted on and the answer is keep the mailing lists. What are the next steps to fixing the mail system which is held together by duct-tape and bailing wire?
It is a seriously large task to get done with everything from someone getting all the packages back into Fedora, architecting how email should flow, getting a documented system and tooling, then dumping the current system, fixing the schema changes, importing the system into a new mail system, and doing that 6 or 7 times to get it actually working. It is a set of tasks needing senior programmers, some DBA time, and senior system administration. It will take about 3 to 6 months of dedicated work to get done. If done on volunteer time, I would put it at 12 months. After that it needs a regular redo and cleanup of code or we will be back where we are currently very fast.
Where we are currently is a system running on old dead code, can't be reinstalled, with a possibly slightly corrupted disk partition (though I think I fixed all of the items) and needing to be coddled regularly to keep going. It has hundreds of thousands of 'held' dead spam messages, similar number of bounces for dead accounts (mailman3 did not implement auto-unsubscribe til after the version we are running) and other oddities.
Or the infrastructure team can work on keeping the build system running which is also fairly touchy with daily restarts, random reboots, etc. That is also more than a full-time job for both the volunteers and dedicated people.
Look I am not wanting to move to discourse, but voting is not going to fix anything.
-- Maxwell G (@gotmax23) Pronouns: He/They _______________________________________________ devel mailing list -- devel@lists.fedoraproject.org To unsubscribe send an email to devel-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
Stephen Smoogen wrote:
So let us say it is voted on and the answer is keep the mailing lists. What are the next steps to fixing the mail system which is held together by duct-tape and bailing wire?
[etc.]
Thanks for confirming that the decision was actually already made elsewhere and that this whole RFC thread is just a farce. Looks like "I propose" in the original post should really read as "I dictate".
Kevin Kofler
Kevin Kofler via devel wrote:
Stephen Smoogen wrote:
So let us say it is voted on and the answer is keep the mailing lists. What are the next steps to fixing the mail system which is held together by duct-tape and bailing wire?
[etc.]
Thanks for confirming that the decision was actually already made elsewhere and that this whole RFC thread is just a farce. Looks like "I propose" in the original post should really read as "I dictate".
PS: The process that Stephen describes reminds me of those rent sharks that deliberately let their beautiful old historical houses rot until they fall apart so badly that the rent sharks are allowed to tear them down and build some ugly new concrete box with unaffordable apartments in their place.
Kevin Kofler
Kevin Kofler via devel devel@lists.fedoraproject.org writes:
Kevin Kofler via devel wrote:
Stephen Smoogen wrote:
So let us say it is voted on and the answer is keep the mailing lists. What are the next steps to fixing the mail system which is held together by duct-tape and bailing wire?
[etc.]
Thanks for confirming that the decision was actually already made elsewhere and that this whole RFC thread is just a farce. Looks like "I propose" in the original post should really read as "I dictate".
PS: The process that Stephen describes reminds me of those rent sharks that deliberately let their beautiful old historical houses rot until they fall apart so badly that the rent sharks are allowed to tear them down and build some ugly new concrete box with unaffordable apartments in their place.
This seems ... disrespectful at best ... toward the people who put their time into keeping the existing email setup working. It seems like a way to make them throw up their hands and find a more rewarding way to spend their time, thus accelerating the demise of this mailing list.
As I've said before, I don't welcome a transition to a siloed web forum at all. But one should not discount the costs of keeping an independent email system working on today's net. It's *not* fun.
Thanks,
jon
On Thu, Apr 27, 2023 at 01:39:00PM +0200, Kevin Kofler via devel wrote:
Kevin Kofler via devel wrote:
Stephen Smoogen wrote:
So let us say it is voted on and the answer is keep the mailing lists. What are the next steps to fixing the mail system which is held together by duct-tape and bailing wire?
[etc.]
Thanks for confirming that the decision was actually already made elsewhere and that this whole RFC thread is just a farce. Looks like "I propose" in the original post should really read as "I dictate".
I... don't understand how you reached that conclusion.
My current understanding is: * Matthew posted about a number of issues / concerns with mailing lists. These concerns are completely 100% unrelated to our current list infrastructure. If we had the very latest mailman3 from upstream running smoothly... it would still be a mailing list and it would really have almost all of the same concerns.
I completely understand where smooge is coming from here. Am I frustrated at the current state of our mailing list infrastructre? Oh so very much. Do I hope we can make it better? Oh so very much.
Are mailing lists going to 'go away' soon? NO. There's a number of things we use lists for that will be difficult to move over to discussion, and that leads me to...
* Matthew said he was going to propose that FESCo decide if moving Changes discussion to discuss was something they wanted to do. I can't see how you think thats been decided? Much less turning off mailing lists. I personally am willing to learn and try it... but I don't even have a good sense where the rest of FESCo is on this. And this is just one part of devel discussions.
Anyhow, I 100% disagree with you that this is 'a farce'. I think it's been useful, I think we have gotten:
* More information on interacting with discussion via email and if it will meet the needs of existing devel folks. * A sense of people who will probibly not want to participate at all, even via an email interface. * Some more use cases mailing lists provide that we should consider a way to provide if we move more things to discussion (example, I think some kind of better/easier public archive for drive by contributors and those that need to look back in history would be good) * Some suggestions for improvements we can ask discourse folks to make to make more people happy.
The next steps here would be for Matthew to ask fesco about moving changes discussions over to discuss and them to vote/consider that. After that, I hope we can fix some of the issues identified here before we do anything further to move more discussions over there.
Does this mean that someday we might turn off mailing lists? Sure, it could, but it could also mean things don't work out on discussion and things never fully move to it, or move, but then move back.
I'll note that we just released Fedora 38... and the users mailing list has had a increase in posts. It had like 60-70 in the last week. The askfedora category in discussion had more than that overnight last night. It's pretty dramatic. Granted this could be due to us advertising that as the way to get help, and users asking questions aren't the same as developers, but still... It's a LOT more interaction.
Anyhow, I'll stop.
kevin
Kevin Fenzi wrote:
I... don't understand how you reached that conclusion.
My current understanding is:
- Matthew posted about a number of issues / concerns with mailing lists.
These concerns are completely 100% unrelated to our current list infrastructure. If we had the very latest mailman3 from upstream running smoothly... it would still be a mailing list and it would really have almost all of the same concerns.
I completely understand where smooge is coming from here.
Well, I see two (almost orthogonal) lines of argumentation here: * Matthew and others have proposed moving off the mailing list due to alleged shortcomings that mailing lists have by design. The issue is that this misses that Discourse has much worse shortcomings by design. (And "by design" is usually synonymous with "unfixable".) * Smooge has brought up that the mailing list has in his view been so badly neglected over the last few years that it will have to be replaced anyway, no matter whether the replacement is actually better or worse.
Am I frustrated at the current state of our mailing list infrastructre? Oh so very much. Do I hope we can make it better? Oh so very much.
I would hope so too.
Anyhow, I 100% disagree with you that this is 'a farce'. I think it's been useful, I think we have gotten:
- More information on interacting with discussion via email and if it
will meet the needs of existing devel folks.
- A sense of people who will probibly not want to participate at all,
even via an email interface.
- Some more use cases mailing lists provide that we should consider a
way to provide if we move more things to discussion (example, I think some kind of better/easier public archive for drive by contributors and those that need to look back in history would be good)
- Some suggestions for improvements we can ask discourse folks to make
to make more people happy.
The next steps here would be for Matthew to ask fesco about moving changes discussions over to discuss and them to vote/consider that. After that, I hope we can fix some of the issues identified here before we do anything further to move more discussions over there.
Do you not see my point here? You have gotten a sizable amount of feedback pointing out showstoppers that make Discourse anywhere from impractical to unusable for several existing contributors, many of which are by design and will never be fixed, or at least require a lot of coding that nobody is signing up for (e.g., an NNTP gateway, plus, that will also suffer from some of the core limitations of the e-mail notification system, e.g., not picking up edits (unless maybe if the gateway "cancels" the post and resends the edited version? But NNTP post cancellation is not universally supported)). Yet, you are still set on pushing this change forward and already discussing the next steps (FESCo vote, "fix[ing] some of the issues" (but you will never be able to fix all of them because several are by design), deployment). That, in my view, makes the RFC thread a farce, because the option that ought to be the default (keep the status quo) appears to not even be under discussion.
I'll note that we just released Fedora 38... and the users mailing list has had a increase in posts. It had like 60-70 in the last week. The askfedora category in discussion had more than that overnight last night. It's pretty dramatic. Granted this could be due to us advertising that as the way to get help, and users asking questions aren't the same as developers, but still... It's a LOT more interaction.
It is certainly both: end users (as opposed to developers) are used to web forums, and of course if you link to the forum everywhere and the mailing list is hidden behind several clicks and has a more complicated signup process, people will choose the forum.
Kevin Kofler
On Fri, Apr 28, 2023 at 06:14:37PM +0200, Kevin Kofler via devel wrote:
Well, I see two (almost orthogonal) lines of argumentation here:
- Matthew and others have proposed moving off the mailing list due to alleged shortcomings that mailing lists have by design. The issue is that this misses that Discourse has much worse shortcomings by design. (And "by design" is usually synonymous with "unfixable".)
I know you maintain this, but I don't agree. I think it has some shortcomings, but the advantages outweight those and we can try and improve those.
- Smooge has brought up that the mailing list has in his view been so badly neglected over the last few years that it will have to be replaced anyway, no matter whether the replacement is actually better or worse.
Well, there's work ongoing to clean up and package the current mailman3 stack, and it's getting closer. I personally see us moving to that so we don't need to deal with that issue and can move things naturally as we want instead of forced by support.
I guess the biggest timebomb is that the current host is rhel7 and that goes end of life next year.
Do you not see my point here? You have gotten a sizable amount of feedback pointing out showstoppers that make Discourse anywhere from impractical to unusable for several existing contributors, many of which are by design and will never be fixed, or at least require a lot of coding that nobody is signing up for (e.g., an NNTP gateway, plus, that will also suffer from some of the core limitations of the e-mail notification system, e.g., not picking up edits (unless maybe if the gateway "cancels" the post and resends the edited version? But NNTP post cancellation is not universally supported)).
I see a number of people who expressed concerns and issues. I hope that they go and look at the current email interface and find it good enough to use. I think a lot of the early posts in this thread had people unaware of how (or indeed if at all) they could interface with discourse without using the web interface. I hope that people are more aware now of the options.
Yet, you are still set on pushing this change forward and already discussing the next steps (FESCo vote, "fix[ing] some of the issues" (but you will never be able to fix all of them because several are by design), deployment). That, in my view, makes the RFC thread a farce, because the option that ought to be the default (keep the status quo) appears to not even be under discussion.
I was simply saying what the next step is. Keeping the status quo is completely an option (at least short term) as FESCo might say "nope, we don't want to do this". Now, if so, I hope there would be a "because we require X, Y, Z" to work on in that case. I also can't speak for anyone but myself, so the council could decide something here, but Matthew clearly said he wanted to try this with changes and would ask FESCo, so I don't see that happening.
Anyhow, I strongly disagree that this is a farce.
kevin
On Wed, 2023-04-26 at 19:27 +0000, Maxwell G wrote:
On Wed Apr 26, 2023 at 18:04 +0200, Vitaly Zaitsev via devel wrote:
On 20/04/2023 23:20, Matthew Miller wrote:
It’s time to transform the Fedora devel list into something new
I think such serious questions should be put to a vote. Not a FESCo vote, but a vote for all Fedora contributors (can be combined with the next FESCo elections).
I think having this as a "ballot referendum" of sorts is a good idea.
A poll like this would have an inherent problem: it's ineffective to have *only* the people who are already in a place vote on whether a measure to get new people into the place is a good idea.
This is the same as the NIMBY problem in municipal politics: if you give the residents of any area too much say over development in that area, they will always tend to oppose it on the grounds that it's bad for *them*. They have no inherent motivation to consider the interests of other people who might want to live or work in the area, but who cannot.
Adam Williamson wrote:
A poll like this would have an inherent problem: it's ineffective to have *only* the people who are already in a place vote on whether a measure to get new people into the place is a good idea.
Yet this approach is working fine for, e.g., Debian.
This is the same as the NIMBY problem in municipal politics: if you give the residents of any area too much say over development in that area, they will always tend to oppose it on the grounds that it's bad for *them*. They have no inherent motivation to consider the interests of other people who might want to live or work in the area, but who cannot.
But the local residents are often precious allies in fighting things such as new highways, construction projects destroying fertile soil, etc. that make things worse for everyone by: accelerating the climate crisis, causing pollution, eating up soil needed for agriculture, etc. So guess what, I often find myself supporting this kind of local initiatives from the other end of the city. Would you be happy if your city builds a huge highway right through your previously quiet neighborhood? Would you find that fair?
Kevin Kofler
-
Adam Williamson -
Alec Leamas -
Aleksandra Fedorova -
Alexander Bokovoy -
Alexander Ploumistos -
Ben Cotton -
Benson Muite -
Björn Persson -
Carl George -
Chris Adams -
Christopher Klooz -
Dan Horák -
Daniel Alley -
Daniel P. Berrangé -
Dominik 'Rathann' Mierzejewski -
Emmanuel Seyman -
Fabio Valentini -
Florian Weimer -
Fulko Hew -
Gabriel Ramirez -
Ian Pilcher -
Jarek Prokop -
Jaroslav Prokop -
Jonathan Corbet -
Josh Boyer -
JT -
Jun Aruga (he / him) -
Jóhann B. Guðmundsson -
Kamil Paral -
Kevin Fenzi -
Kevin Kofler -
Kevin P. Fleming -
Mario Torre -
Marius Schwarz -
Mark Wielaard -
Matthew Miller -
Mattia Verga -
Maxwell G -
Michael Catanzaro -
Michael J Gruber -
Miro Hrončok -
Miroslav Suchý -
Nils Philippsen -
Panu Matilainen -
Peter Boy -
Ralf Corsépius -
Randy Barlow -
Richard Shaw -
Richard W.M. Jones -
Robert Marcano -
Roberto Ragusa -
Sam Varshavchik -
Sandro -
Simo Sorce -
Solomon Peachy -
stan -
Stephen Smoogen -
Todd Zullinger -
Tom Hughes -
Tomi Lähteenmäki -
Vitaly Zaitsev -
Vít Ondruch -
Zbigniew Jędrzejewski-Szmek