The Problems of Perl: The Future of Bugzilla

[Max Kanat-Alexander]

On Thu, 03 May 2007 17:43:51 +0100 Nick Barnes <Nick.Barnes at pobox.com>
wrote:
> I broadly agree with his views about Perl the language, although the
> quality of the Perl in Bugzilla has improved enormously since I first
> worked on it in the 2.10 days.

	I'm glad you think so! :-) I certainly agree. It took a lot of
work to get there, though--the language doesn't really help out.

> If one were determined to accomplish this, I guess one would start by
> choosing a Python templating system (e.g. Cheetah).

	I'm going to look at web frameworks in various languages,
probably starting with Pylons in Python and Rails in Ruby.

	The idea is to prototype some small part of Bugzilla in that
framework to see how it would work. Say--just enter_bug.cgi and
post_bug.cgi.

>  It looks to me as
> if converting TT templates into Cheetah could be done in a
> semi-automated way.

	I think this is actually true for most templating systems, for
our simpler templates. Some would need to be re-written by hand, but
they'd probably get simpler.

> However, it's a long and uncomfortable path and it's not clear whether
> it's worth the effort.  If all I wanted was a system similar to
> Bugzilla but in a usable language, I would start from scratch.

	And that's what many people have done. I like Bugzilla, though.

> Also, I have a vested interest in Bugzilla remaining Perl: it makes me
> money in two different ways.  Firstly, customizing Bugzilla directly
> in Perl, which clients are generally unwilling to attempt for
> themselves. Secondly, providing custom interfaces to Bugzilla in
> Python, in which clients often are willing to work themselves.

	I think that's an excellent demonstration of why it *should* be
transitioned to another language. :-)

	-Max
-- 
http://www.everythingsolved.com/
Competent, Friendly Bugzilla Services. And Everything Else, too.