Philipp Kern

Do you think the solution is to pretend bugs exist no longer just so that we can release? I'd prefer fixing bugs and not releasing.

> Fellow developers, do you really want to release? Is the Release Team expected to fix all the bugs you file?

Are these bugs being reported because the bugs never existed before? Or are they only now being reported? A bug report doesn't cause the bug to exist, it reports its existence. The actual bug count is unknown, and reporting bugs helps bring our picture of reality closer to the truth.

While it would be good if Debcamp resulted in bugs being resolved, its still preferable for the bugs to at least be reported than not. I would prefer to have a more accurate bug count than to release ignorant of all the bugs being found.

> It is a shame that the bug count for Lenny is only rising since the beginning of Debcamp (about 2008-08-02).

"only rising"? I see rises and falls over that period. Especially since 2008-08-10, a significant trend downward is noticeable.

Is it valid to extrapolate from that? Perhaps, perhaps not. If not, though, then why extrapolate only from the rise you observe?

Hi Ben,

actually my post was intended not in the way Madduck and you read it, but well... Of course bugs need to be reported. The point is that very few people concentrated on bug fixing on Debcamp and Debconf. Almost nobody is really interested in actually making the release happen. Which is what I was annoyed about. Well, maybe I should do a followup post then.

To your observation about the significant trend downwards. That was basically me removing packages from testing, lowering severities, uploading patches and scheduling binNMUs for a bug in gdc-4.1 fixed by dondelelcaro, which closed 10 RC bugs at once. So maybe it's not entirely reasonable to draw too much of conclusions based on that.

The real problem is that people mostly only care about their own packages, may they be pet packages or not. We regularily get freeze exception requests just because there's a new upstream version. People don't look over their plate (now that may be a German term) and look what's going on with other, negclected packages. (Although packages maintained by QA look like they are in a better shape, than packages from MIA maints... etc.)

Nightly greetings from Argentina,
Philipp Kern

PS: It's actually pretty annoying if people insert wrong email addresses if you actually want to reply. Discovering that you already have a mail drafted entirely but that there is no mail address to send it to.