Another idea, what good and bad would come from us upgrading to the 1.6 beta?
-Mike
On Dec 20, 2007 7:20 PM, Mike McGrath mmcgrath@redhat.com wrote:
Another idea, what good and bad would come from us upgrading to the 1.6 beta?
Well my questions would be:
What are we trying to accomplish? Who is on base to accomplish this? I am not sure that this would be the best 'python' learning experience but since I just made our PFY do moin as his first project.. I should probably do the same.
Who are the human chattel we have to throw at it? When do we need to see results of what we are trying to accomplish?
Mike McGrath wrote:
Another idea, what good and bad would come from us upgrading to the 1.6 beta?
What features were you thinking it provides us with? Better ability to cache pages?
Here's another thought: is it only page saves which are causing us problems? I think moin has a similar URL for all actions that cause writing to the disk. For instance, saving an edit appears to call PAGENAME#preview and starting an edit uses PAGENAME?action=edit. Could we cluster just requests to read from the wiki and filter writes to a single app server by using these URLs?
-Toshio
Toshio Kuratomi wrote:
Mike McGrath wrote:
Another idea, what good and bad would come from us upgrading to the 1.6 beta?
What features were you thinking it provides us with? Better ability to cache pages?
That, the other thing I'm thinking of is if we're going to start patching our instance heavily, perhaps we want to stick with 1.6. I don't necessarily want to upgrade to 1.6 but I do want us to talk about it.
Here's another thought: is it only page saves which are causing us problems? I think moin has a similar URL for all actions that cause writing to the disk. For instance, saving an edit appears to call PAGENAME#preview and starting an edit uses PAGENAME?action=edit. Could we cluster just requests to read from the wiki and filter writes to a single app server by using these URLs?
The page saves are causing our inability to cluster Moin in any efficient way. I've thought about altering proxy location by action, that would help with some things but I've yet to set it up and see if it works the way we think it will. (for example, altering the user, watched pages, etc are still a question mark in my head).
-Mike
Mike McGrath wrote:
Toshio Kuratomi wrote:
Mike McGrath wrote:
Another idea, what good and bad would come from us upgrading to the 1.6 beta?
What features were you thinking it provides us with? Better ability to cache pages?
That, the other thing I'm thinking of is if we're going to start patching our instance heavily, perhaps we want to stick with 1.6. I don't necessarily want to upgrade to 1.6 but I do want us to talk about it.
Very good point. That would definitely be another reason to move to 1.6 so we're closer to the code base that our patches are going to apply to upstream.
Here's another thought: is it only page saves which are causing us problems? I think moin has a similar URL for all actions that cause writing to the disk. For instance, saving an edit appears to call PAGENAME#preview and starting an edit uses PAGENAME?action=edit. Could we cluster just requests to read from the wiki and filter writes to a single app server by using these URLs?
The page saves are causing our inability to cluster Moin in any efficient way. I've thought about altering proxy location by action, that would help with some things but I've yet to set it up and see if it works the way we think it will. (for example, altering the user, watched pages, etc are still a question mark in my head).
It looks like subscribing to a page and starting the user preferences editor via the link in the sidebar can be grabbed via the ?action= string as well. Saving of the User Preferences page looks like it doesn't use a special URL but perhaps sending all POST requests to the writable server would catch that.
-Toshio
Mike McGrath kirjoitti:
Another idea, what good and bad would come from us upgrading to the 1.6 beta?
Good: Xapian integration and with it, hopefully a better search? http://moinmo.in/HelpOnXapian
Ville-Pekka Vainio wrote:
Mike McGrath kirjoitti:
Another idea, what good and bad would come from us upgrading to the 1.6 beta?
Good: Xapian integration and with it, hopefully a better search? http://moinmo.in/HelpOnXapian
Yeah, we can certainly do that too.
I guess I started this thread because I see all the changes / patches and things that could possible be implemented into our wiki and I'm wondering if sticking with upstreams next release would help mitigate any integration issues we might have. It just feels like a lot of work is being done (good) and I want to make sure we're in the best possible spot for our future.
-Mike
Hi all,
Just a reminder, that Moin 1.6.0 is released: http://moinmo.in/MoinMoinRelease1.6
Apparently 1.6 also has discussion pages, I've never tested those, but it seems this is something people have been missing.
On Tue, 25 Dec 2007, Ville-Pekka Vainio wrote:
Hi all,
Just a reminder, that Moin 1.6.0 is released: http://moinmo.in/MoinMoinRelease1.6
Apparently 1.6 also has discussion pages, I've never tested those, but it seems this is something people have been missing.
Interesting. We'll have to work with the Moin maintainer and see if we can get this up and going "as is".
Those of you working on Moin, how much will this mess up what you're working on, is anyone against upgrading to 1.6? Is anyone particularly for it?
-Mike
Mike McGrath wrote:
On Tue, 25 Dec 2007, Ville-Pekka Vainio wrote:
Hi all,
Just a reminder, that Moin 1.6.0 is released: http://moinmo.in/MoinMoinRelease1.6
Apparently 1.6 also has discussion pages, I've never tested those, but it seems this is something people have been missing.
Interesting. We'll have to work with the Moin maintainer and see if we can get this up and going "as is".
Those of you working on Moin, how much will this mess up what you're working on, is anyone against upgrading to 1.6? Is anyone particularly for it?
I'm for it.
I handed off my patch to MrBawb but we knew we'd have to maintain a version of the patch against 1.5.x for our use and one against upstream's development branch. Having our patch apply to something more recent can only help that situation. MrBawb, do you concur?
-Toshio
On Tue, 25 Dec 2007, Toshio Kuratomi wrote:
I'm for it.
I handed off my patch to MrBawb but we knew we'd have to maintain a version of the patch against 1.5.x for our use and one against upstream's development branch. Having our patch apply to something more recent can only help that situation. MrBawb, do you concur?
Yeah, I expect that the Moin developers will want to only apply the patch to their development tree. They were supposed to look over the code and get back to me about that, but I haven't heard from them. I'll ping them again.
infrastructure@lists.fedoraproject.org