Greetings.
So I've been wondering how we can measure usage of Fedora vs unofficial remixes that have removed the Fedora Trademarks and while speaking with Seth regarding one of the ideas I had on how we could potentially implement this he pointed it out to me that we might be legally prohibited to gather this information with or without explicit user consent and forwarded me to you guys to straighten this out.
So my question to you guys is.
Is there anything legally prohibiting or restricting us on verifying and gather information if users are running Fedora or are running some unofficial remix that have removed the Fedora trademarks?
If the answer is yes.
What are we allowed to do?
What are we not allowed to do?
If we are not allowed to gather this data in anyway or form can the unofficial remix be legally bound to notify us of their existence and or provide us with that data so we could have some kind of an idea on how many remixes are out there which Fedora is "upstream" for?
Regards JBG
On 10/29/2010 09:57 AM, "Jóhann B. Guðmundsson" wrote:
Is there anything legally prohibiting or restricting us on verifying and gather information if users are running Fedora or are running some unofficial remix that have removed the Fedora trademarks?
It is difficult/impossible to answer this question without specifics on implementation. This may seem a bit opposite from what you've asked for, but it is how lawyers prefer to work.
Let me instead say the following:
Smolt already records the OS/Distro when users opt-in. I assume it looks in /etc/redhat-release or /etc/fedora-release (and other places for non-Fedora/Red Hat distros).
It is fine to look at the Smolt counts to guestimate how many remixes are out there, with the understanding that some remixes may not be using Smolt.
If you have another specific proposal, I'd be willing to look it over.
If we are not allowed to gather this data in anyway or form can the unofficial remix be legally bound to notify us of their existence and or provide us with that data so we could have some kind of an idea on how many remixes are out there which Fedora is "upstream" for?
I am unaware of any existing requirement that remixes "notify" anyone of their existence. It is likely that if we attempted to impose such a requirement upon modification of Fedora, it would conflict with the Free Software licensing that Fedora is built on. FWIW, I do not see the merit of any such requirement, except to make it more difficult for remixes to exist. Unless a remix is using the Fedora trademarks outside of the Trademark Usage Guidelines, my inclination is to simply let them do what they wish.
~spot
On 10/30/2010 12:52 PM, Tom "spot" Callaway wrote:
It is difficult/impossible to answer this question without specifics on implementation. This may seem a bit opposite from what you've asked for, but it is how lawyers prefer to work.
Via updates was what I was thinking either a check or setting up Fedora/Generic/<insert registered remix> Channel something similar to Red Hat channel basically.
Preferable a solution the user would not have to opt in about... ( Those only give the half of the picture at best )
The end goal here simply for us being able to measure and gather statistic for various purposes on Fedora vs Unofficial remixes.
As things stand now we cant even say how many remixes are out there nor how much of Fedora's overall stats and success belongs to them thus for example no way for us to do some research in what they are doing right and we are doing wrong or how much these remixes contribute back to Fedora and so on and so fourth simply because we have no way of telling that.
JBG
On 10/30/2010 10:47 AM, "Jóhann B. Guðmundsson" wrote:
Via updates was what I was thinking either a check or setting up Fedora/Generic/<insert registered remix> Channel something similar to Red Hat channel basically.
Preferable a solution the user would not have to opt in about... ( Those only give the half of the picture at best )
The difficulty in forced tracking is that we have to be very very very careful about what information we gather in such situations. With opt-in scenarios, it is notably easier.
The end goal here simply for us being able to measure and gather statistic for various purposes on Fedora vs Unofficial remixes.
Okay, I can perhaps see some intellectual value in this, but I would point out that I do not think it is a notable weakness if we are unable to have exact statistics on usage and generation of remixes. Smolt should give us a rough count on those that are in the wild (and using Smolt), which is probably sufficient for these "various purposes".
As things stand now we cant even say how many remixes are out there nor how much of Fedora's overall stats and success belongs to them thus for example no way for us to do some research in what they are doing right and we are doing wrong or how much these remixes contribute back to Fedora and so on and so fourth simply because we have no way of telling that.
Well, the requirement that Remixes use either the Secondary Mark ("Fedora Remix") or none of the Fedora trademarks means that we can generally avoid the problem of confusing remixes with the "real" Fedora in tracking efforts such as Smolt.
If the initiative is truly to determine what the remixes do right/wrong, raw statistics will not help in that effort, only involvement and interaction with the communities making those remixes will.
We can encourage Fedora Remixes to list themselves on a wiki page, but I do not think we need to require this. I do not think it is in the interest of anyone to make it any more difficult for someone to make a derivative work or remix of Fedora.
~spot
On Mon, Nov 01, 2010 at 09:00:41AM -0400, Tom spot Callaway wrote:
We can encourage Fedora Remixes to list themselves on a wiki page, but I do not think we need to require this. I do not think it is in the interest of anyone to make it any more difficult for someone to make a derivative work or remix of Fedora.
I could see a web application that made it easy for remixers to provide all sorts of information about what, where, how, why, who, when, and so forth. Even more, I could see an active special interest group in Fedora that attracted remixers by giving them reasons to be involved. A sort-of meta-community of remixers, who themselves might be the tip of many thousands of remix users.
The concern is, anything that is automatic without opt-in is going to need a HUGE amount of visibility to the Fedora community. People are wary of being tracked and reported on, so they would have to know early what was being tracked, why, how, how it was aggregated, anonymized, and so forth. All of that decided, openly, before a single line of code need be written.
Smolt was created originally as a response to a situation where we weren't going to go through that last paragraph of hassle with the Fedora community. Could the data needs you are talking about utilize the existing Smolt + grow a remixer community?
- Karsten
On Mon, Nov 01, 2010 at 10:14:22AM -0700, Karsten Wade wrote:
On Mon, Nov 01, 2010 at 09:00:41AM -0400, Tom spot Callaway wrote:
We can encourage Fedora Remixes to list themselves on a wiki page, but I do not think we need to require this. I do not think it is in the interest of anyone to make it any more difficult for someone to make a derivative work or remix of Fedora.
I could see a web application that made it easy for remixers to provide all sorts of information about what, where, how, why, who, when, and so forth. Even more, I could see an active special interest group in Fedora that attracted remixers by giving them reasons to be involved. A sort-of meta-community of remixers, who themselves might be the tip of many thousands of remix users.
The concern is, anything that is automatic without opt-in is going to need a HUGE amount of visibility to the Fedora community. People are wary of being tracked and reported on, so they would have to know early what was being tracked, why, how, how it was aggregated, anonymized, and so forth. All of that decided, openly, before a single line of code need be written.
[...snip...]
In fact, opt-out solutions may be expressly forbidden in some locations, or may require complex contortions of shrink-wrapped agreements that we're loathe to create in Fedora.
On 11/01/2010 05:14 PM, Karsten Wade wrote:
On Mon, Nov 01, 2010 at 09:00:41AM -0400, Tom spot Callaway wrote:
We can encourage Fedora Remixes to list themselves on a wiki page, but I do not think we need to require this. I do not think it is in the interest of anyone to make it any more difficult for someone to make a derivative work or remix of Fedora.
I could see a web application that made it easy for remixers to provide all sorts of information about what, where, how, why, who, when, and so forth. Even more, I could see an active special interest group in Fedora that attracted remixers by giving them reasons to be involved. A sort-of meta-community of remixers, who themselves might be the tip of many thousands of remix users.
We have had http://fedoraproject.org/wiki/DerivedDistributions which is not working out that well I would say. Perhaps as you suggest a stand alone web application would work better than the wiki but I still fear that it stays the same since there is no motivating factor or direct requirements for them to provide that information to us thus it would not be maintained any better then a simple wiki page.
If you think about it they already get what they need from us why would they contribute anything back if they dont have to?
I personally see nothing wrong in having them register those remixes with us for measurements and research purposes at least we would be getting something back from them that way instead of those remixes just remaining unknown entity out there without us knowing of their existence nor if they are actually contributing anything back to the project or directly upstream for that matter ( upstream being other than Fedora in that term ).
Regarding Remix SIG I think it would be better to strengthen the already existing Spin SIG rather then creating a new one...
The concern is, anything that is automatic without opt-in is going to need a HUGE amount of visibility to the Fedora community. People are wary of being tracked and reported on, so they would have to know early what was being tracked, why, how, how it was aggregated, anonymized, and so forth. All of that decided, openly, before a single line of code need be written.
Certainly.
Smolt was created originally as a response to a situation where we weren't going to go through that last paragraph of hassle with the Fedora community. Could the data needs you are talking about utilize the existing Smolt + grow a remixer community?
Comparing what's present on smolt with what's mentioned on distrowatch then we only got a handful of potential derivatives that are being reported and there are only 3 that are present in smolt, on distrowatch and on our wiki page at the same time ( Ojuba, Mythdora and Linpus).
Without knowing what these remixes are how are we going to be able to tell good remixes from potentially bad ones?
As an example smolt is reporting 25 people using Fedora release 15 ( Finian ). What we know is that Fedora 15 is in development and is referred by us and smolt as rawhide and certainly has not been officially released yet.
Finian is not one of the names that got suggested for Fedora release 15 names. ( That name should be announced today given that the voting ended at midnight ).
So is Fedora release 15 ( Finian ) a cracked version of Fedora that some 25 unsuspecting users are using thinking that it is the latest and greatest from Fedora or is it simply Rawhide remixed in Finnish?
JBG
2010/11/2 "Jóhann B. Guðmundsson" johannbg@gmail.com:
Finian is not one of the names that got suggested for Fedora release 15 names. ( That name should be announced today given that the voting ended at midnight ).
So is Fedora release 15 ( Finian ) a cracked version of Fedora that some 25 unsuspecting users are using thinking that it is the latest and greatest from Fedora or is it simply Rawhide remixed in Finnish?
When Jesse Keating's son was born, I temporarily renamed the rawhide release to Finian (his son's name) as a way of recognizing Jesse for the hard work he's put into the release engineering team. A few weeks later, we reverted back to calling it "Rawhide". So there's no need to worry about Finian.
-- Jared Smith Fedora Project
On 11/02/2010 11:27 AM, Jared K. Smith wrote:
2010/11/2 "Jóhann B. Guðmundsson"johannbg@gmail.com:
Finian is not one of the names that got suggested for Fedora release 15 names. ( That name should be announced today given that the voting ended at midnight ).
So is Fedora release 15 ( Finian ) a cracked version of Fedora that some 25 unsuspecting users are using thinking that it is the latest and greatest from Fedora or is it simply Rawhide remixed in Finnish?
When Jesse Keating's son was born, I temporarily renamed the rawhide release to Finian (his son's name) as a way of recognizing Jesse for the hard work he's put into the release engineering team. A few weeks later, we reverted back to calling it "Rawhide". So there's no need to worry about Finian.
I was only using it as an example but never the less it was a great way to honour Jesse for his valuable work.
JBG
On Tue, Nov 02, 2010 at 15:57:13 +0000, ""Jóhann B. Guðmundsson"" johannbg@gmail.com wrote:
On 11/02/2010 11:27 AM, Jared K. Smith wrote:
2010/11/2 "Jóhann B. Guðmundsson"johannbg@gmail.com:
Finian is not one of the names that got suggested for Fedora release 15 names. ( That name should be announced today given that the voting ended at midnight ).
So is Fedora release 15 ( Finian ) a cracked version of Fedora that some 25 unsuspecting users are using thinking that it is the latest and greatest from Fedora or is it simply Rawhide remixed in Finnish?
When Jesse Keating's son was born, I temporarily renamed the rawhide release to Finian (his son's name) as a way of recognizing Jesse for the hard work he's put into the release engineering team. A few weeks later, we reverted back to calling it "Rawhide". So there's no need to worry about Finian.
I was only using it as an example but never the less it was a great way to honour Jesse for his valuable work.
Maybe we can name F19 Ox13?
-
Bruno Wolff III -
Jared K. Smith -
Jóhann B. Guðmundsson -
Karsten Wade -
Paul W. Frields -
Tom Callaway