A reporter shouldn't be able to confirm his own bug

Sometimes I report a bug and then start fixing it myself, in which case it goes straight from "Unconfirmed" to "In Progress". If I stop working on it, should I still not be allowed to set it back to "Confirmed"? Maybe it's reasonable to prevent the reporter from moving a bug report from Unconfirmed to Confirmed, while still letting them change it from anything else to Confirmed. That way we discourage bad practice (confirming your own bug reports) while not actually preventing it.

Some people (like myself) is in a position that they can report a bug and confirm it. I do it daily with oops bugs.

Is that such a bad practice? Does it really hurt?

The idea in my mind is that it's ripe for abuse, like triaging, where ubuntu-qa membership is a prerequisite. Maybe we could apply the same logic to confirming one's own bugs.

On Fri, Sep 01, 2006 at 07:06:09PM -0000, Rocco Stanzione wrote:
> The idea in my mind is that it's ripe for abuse, like triaging, where
> ubuntu-qa membership is a prerequisite. Maybe we could apply the same
> logic to confirming one's own bugs.

I agree that Importance is something that only selected individuals should be
able to change. That prevents the reporter changing the importance
trying to "FIX ME FIX ME FIX ME! I'M A IMPORTANT BUG" thing.

I don't see lots of people changing the bug status trying to get more
attention or abusing the system. Do you see this kind of behaviour in Ubuntu
bugs?

--
Diogo M. Matsubara

I wouldn't call it an epidemic, but I've certainly seen bugs marked as confirmed by their reporters with no comments from anyone else. As far as triaging (this would be a new ticket), maybe we'd get the best of the getting-help-with-triaging and cutting-down-on-abuse worlds if the must-be-a-qa requirement applied only to triaging your own bugs.

I've seen many bugs submitted by email that confirmed with the same email.

An example (not wishing to pick on anyone) is bug #65322

affects distros/ubuntu/subterfugue
status confirmed
subscribe motu
-------
This strikes me as wrong as you are saying your script is infalliable. In practice the bug squad checked every one and marked some down as needs info when we couldn't verify.

I've been told that with sync requests the submitter needs to confirm (although I'm not sure why) when I raised this on the bug squad email list.

Caroline, sometimes Launchpad e-mail notifications include things that haven't happened yet (bug 41063). So not *all* bugs that look like they were confirmed immediately actually were.