Kernel Summit 2006: Kernel quality and development process [LWN.net]

Kernel Summit 2006: Kernel quality and development process

Posted Jul 18, 2006 9:53 UTC (Tue) by job (guest, #670)
Parent article: Kernel Summit 2006: Kernel quality and development process

I've never really liked the Bugzillas. Too much "database" and too little "conversation".

If you go to a mailing list with your problem and attract attention, you may get intelligent questions asked back that not only gets more information but elevates your own understanding of the problem (which may not be that great which is why you came for help in the first place).

Contrast that with a Bugzilla where when you report something you get "dont' understand, can't reproduce, wontfix". Sure, nothing stops you from having an intelligent discussion on a Bugzilla, but nobody really does. It's not designed for that.

I'm not saying the tool is useless, it's great to track bugs, but there's tremendous room for improvement.

(Log in to post comments)

Kernel Summit 2006: Kernel quality and development process

Posted Jul 18, 2006 13:49 UTC (Tue) by seyman (subscriber, #1172) [Link]

> I've never really liked the Bugzillas. Too much "database" and too little "conversation".

This seems to depend on the developpers using said Bugzillas and not on the tool per se.

> I'm not saying the tool is useless, it's great to track bugs, but there's tremendous room for improvement.

If you're serious about this, please come talk to the Bugzilla devs on IRC or via the mailing list (details are on http://www.bugzilla.org/). Better yes, file bugs in bugzilla.mozilla.org against the Bugzilla product.

Kernel Summit 2006: Kernel quality and development process

Posted Jul 30, 2006 18:01 UTC (Sun) by rmk (subscriber, #7713) [Link]

> This seems to depend on the developpers using said Bugzillas and not on the tool per se.

This is not my experience. If I look at my bugs in the kernel bugzilla, there are two outstanding bugs at the time of writing - 6716 and 6815.

6716 is presumed to be solved in 2.6.18-rc2, but asking the submitter to check this resulted in no response.

6815 currently defies logical explaination and the submitter also seems to have one to ground.

You claim that it depends on developers - I disagree. My experience has been that it depends more on users responding to developers, rather than dumping a bug in and running away. The submitter of a bug needs to takes on a certain responsibility to assist the resolution of the bug. Without that, developers are left out in the cold and have no option but to throw away potentially valid bug reports.

Because of that, I dislike bugzilla intensely. The tool itself is fine, but the effect it has (that I've perceived so far) is far from desirable.

Kernel Summit 2006: Kernel quality and development process

Posted Jul 18, 2006 16:45 UTC (Tue) by zooko (subscriber, #2589) [Link]

I used trac as a forum for conversation (at my workplace). I used to use the Debian bug tracking system (by e-mail). I've never successfully had a conversation on bugzilla. Possibly a coincidence, or possibly a flaw in the design of bugzilla.