Re: Bugzilla for GNOME (was Re: GNOME CVS: nautilus mathieu)

[Christopher James Lahey <clahey helixcode com>]

On 27 Oct 2000, Martin Baulig wrote:

> Joe Shaw <joe helixcode com> writes:
> 
> > A debbugs email wrapper needs to be written OR an email interface needs to
> > be installed and have support added to bug-buddy.
> 
> Adding Bugzilla support to bug-buddy is much better than an email interface;
> this'll also allow us to add some ideas from
> http://bugzilla.eazel.com/helper.html.

I disagree here.  Implementing bug-buddy support as an email interface
allows you to implement whatever ideas you want from helper.html while
also allowing easy offline bug reports.  Getting data from the client to a
server offline is a solved problem.  sendmail and company solve this
problem well for email.  Reimplementing this ourselves would be a bad
idea.

> > Also, we should consider whether to move the bugs from bugs.g.o over or
> > whether to drop all of them on the floor.
> 
> I think it's enough if the maintainers of all modules have a look over the
> bugs for their module and resubmit the good ones to Bugzilla, or at least
> give them some time to close and remove all the crap (while debbugs is already
> rejecting new bug reports) before we move.

I'm not convinced about this.  There are a lot of modules in
bugs.gnome.org whose maintainers aren't active but which may become alive
again later.

You said in another main that it's a waste to move all those bugs just to
delete most of them, but this doesn't seem right.  It's much easier to get
rid of bugs than it is to resubmit them.  It's especially easy with
bugzilla.  However, resubmitting bugs is a pain.  You have to copy lots of
data, and you're likely to lose some of it.

Perhaps a good compromise would be to not copy the database, but
a) provide a tool to migrate bugs.  An optimal solution would be to add a
link on bugs.gnome.org bugs pages that migrates the bug to bugzilla when
you click it.
b) Keep the old database around.  There's absolutely no reason to delete
the old database.  I certainly agree we should freeze it, but I see no
reason to remove it.

Thanks,
   Chris