Re: Debbugs

[Don Armstrong <don@donarmstrong.com>]

On Wed, 01 Dec 2010, Ian Jackson wrote:
> We at Xen.org (Xen upstream) are thinking of deploying a debbugs
> instance.
> 
> I found the version in experimental which looks reasonable on first
> glance and Colin Watson told me to look here for the current bzr
> branch: http://bugs.debian.org/debbugs-source/mainline/
> 
> I have a couple of questions if you wouldn't mind:
> 
>  * Would you recommend new deployments use 2.4.2~exp1, the bzr clone,
>    or something else ?

I'd actually use the bzr clone; that's what we're using for Debian
itself. (You can check out the configuration on
bugs-master.debian.org)

>  * Would you be willing to accept patches which arrange that a
>    debbugs instance is always able to serve up its own
>    actually-running source code ?

I'd certainly consider it. I think the code should at least serve up
the upstream, and be easily modifiable to serve up the currently
running instance. [I could see cases where you'd modify debbugs and
didn't want it serving the codebase, though; it's sometimes easier to
just modify the code directly instead of putting in configuration
bits.]

>  * We're probably going to put a somewhat strange filtering script
>    in front of debbugs and our own xen-devel mailing list, to allow
>    us to retrospectively turn mailing list threads into bugs. Do you
>    have any interest in this and/or are you likely to have opinions
>    ? If you like I can send you a copy of my draft spec when I've
>    written it.

Sounds reasonable. A long term goal is to make more of the code
amenable to such filtering scripts plugging in at a lower level, so
they don't have to generate "special" looking messages. [Basically,
you'd do something like:

my $nnn = Debbugs::???::new_bug(message => $message,
          	                package => 'blah',
                                other_pseudoheaders => ....)

and then Debbugs::???::append_to_log(bug=>$nnn, message => $message);]



Don Armstrong

-- 
The computer allows you to make mistakes faster than any other
invention, with the possible exception of handguns and tequila
 -- Mitch Ratcliffe

http://www.donarmstrong.com              http://rzlab.ucr.edu