On Wednesday 17 October 2007 01:53:59 Toshio Kuratomi wrote:
> It'd require patches to moin. IE: We'd either have to
get rid of all of
> the users or gut the mailing section of the wiki I guess. Just a
> reminder to those that aren't familiar with it. Moin iterates over
> every user who has page watch lists (currently in the thousands) to find
> out who to notify. As the number of our users increases and as the
> number of the pages they watch increase moin gets slower and slower on
> page saves.
It would be doable though. When I looked ~ a year ago, there were only
two places in the code that would need to be hacked to replace the email
subsystem. I imagine we could find a single point in the code to hack
if we just made finding the subscribed pages function return
'fedora-wiki-commits-list(a)r.c' instead of parsing each user's config
file and returning that list.
I think I've suggested this before, but shouldn't we first try to fix Moin
rather than spend time wondering if we can convert all the data to Mediawiki?
We have a lot of talented Python people in this project, it shouldn't be
impossible to do.
I do have some of my own interests here, my Summer Code Finland 2007 project
[1] for Fedora was based on Moin, and if Fedora ditched Moin that would
seriously threaten the future of that code. There was also a Fedora GSoC
project on Moin last year, so even though none of this code is yet merged
upstream, a lot of work has been done on Moin by Fedora people.
I don't want to step on anyone's toes here, I do understand the situation and
I see that it causes a lot of trouble for us. I'm just trying to offer some
pro-Moin points here as well.
[1]
http://fedoraproject.org/wiki/SummerOfCode/2007/VillePekkaVainio
--
Ville-Pekka Vainio
vpivaini(a)cs.helsinki.fi