[gnome-love] bugzilla status- volunteers needed to do some dirty, dirty work

[Luis Villa <louie ximian com>]

Bugzilla is badly in need of an upgrade. Has been for ages, of course.
Since roughly the dawn of GNOME time, in fact.[0] Last fall some
progress was made; you can see results at
http://bugzilla-test.gnome.org/ 

The original work was excellent. Unfortunately, for a number of reasons,
the ball has mainly been dropped, and progress has stalled. Someone in
IRC asked tonight 'what can I do to help' so I thought I'd write up an
email- I'm completely without time to do any of this myself. :/ 

There are several open issues, discussed in some more detail here:

http://cvs.gnome.org/lxr/source/bugzilla-new/TODO.gnome

There are two key issues:

The worst one is that there needs to be a bug-buddy->bugzilla bridge,
and it needs to be installed and tested without breaking the current
bug-buddy->bugzilla stuff. There are a couple options here: (1) port the
old bridge to bugzilla 2.18 or (2) write a new bridge. The current code
takes bug-buddy output, turns it into base-64 encoded bugzilla XML, then
send the XML to a (hacked to deal with base-64) bugzilla XML importer.
Ideally a new version would either bypass the bugzilla XML importer or
not require it to be hacked. The TODO file linked above discussed this
all in some detail.

The second 'key' issue is the issue of custom fields. When we initially
ported our stuff to 2.16, we used a bugzilla patch that we thought was
going to be in 2.18 to generate custom fields for the OS Details field,
the OS Versions field, and the GNOME Version field. This patch no longer
appears to be in favor upstream. So... we probably need to rethink how
we've done custom fields, and if possible, simplify them so that they'll
be easier to port to 2.18 when that is available. 

So... hopefully some folks can find some time and think about solving
these problems. It would be a great help to all of gnome if we could
finally get ourselves onto a newer, cleaner bugzilla codebase and work
seriously to reduce the obstacles to future upgrades. 

Luis

[0]Note that the dawn of GNOME time for me was early 2002. YMMV. :)