[kepler-dev] bugzilla and kepler

[Christopher Hylands Brooks]

BTW - I've had a certain amount of bug handling experience:

    - I was bug handler for a commercial company for awhile.
    - I've tried various bug handling systems with the Ptolemy project
    and Gigascale Silicon Research Center including gnats, Bugzilla
    and Keystone.

I thought I'd pass on a few comments. 


I think Bugzilla is a good choice.  We chose Keystone because it was
easier to integrate into the authentication scheme on the website.
If I was choosing again, I would try Bugzilla again.
       
In looking at bug handlers, what it comes down to is that one person
or group of people need to keep the bug database up to date, which
means go through the database and reprioritizing or closing bugs.

Also, the developers need to get used to the idea of submitting good
bug reports.

There are short term bugs (such and such no-longer builds because of a
      typo) - usually not worth a bug report.
Longer term bugs (the software does not do as expected/designed)
      - usually worth a bug report.
Requests for enhancements (RFE)s (It would be cool if)
      - usually worth a bug report.

Bug submitters need training in what constitutes a good bug report.
To me, information about the OS, versions of software are critical.
Bug submitters need to provide detailed step-by-step instructions
on how to reproduce the bug, especially with UI bugs.


Bugzilla is pretty cool, the two issues I have:
1) Authentication of users is wide open.  If a user uses a password
for bugzilla that is the same as another critical password, then
they could be in trouble.

2) A lack of a good email interface.  I would _really _ like to be
able to send email to an alias with the appropriate Subject line,
like Re: [kepler-bug 666] and have the message get appended.

These two issues are probably addressed somewhere in bugzilla.

Having a bug list is very valuable when deciding what to do in the
next release, or when to ship.  I'm looking forward to seeing
how it works out for Kepler.

-Christopher