A chilly week

on

It’s finally become properly autumnal, in the real world and in Debian. One week ago, I announced (on behalf of the whole release team) that Debian 8 “Jessie” had successfully frozen on time.

At 18:00 that evening we had 310 release critical bugs – that is, the number that we must reduce to 0 before the release is ready. How does that number look now?

Well, there are now 315 bugs affecting Jessie, at various stages of progression. That sounds like it’s going in the wrong direction, but considering that over a hundred new bugs were filed just 8 hours after the freeze announcement, things are actually looking pretty good.

Out of those 315 bugs, 91 have been fixed and the packages affected have already been unblocked by the release team. The fixed packages will migrate to Jessie in the next few days, if they continue to be bug-free.

Thirty-four bugs are apparently fixed in unstable but are not cleared for migration yet. That means that the release team has not spotted the fix, or nobody has told us, or the fixed package is unsuitable for some other reason (like unrelated changes in the same upload). You can help by trying to find out which reason applies, and talking to us about it. Most likely nobody has asked us to unblock it yet.

Speaking of unblocks, we currently have twenty-four requests that need to be looked at, and a further 20 which are awaiting more information from the maintainer. We already investigated and resolved 260 requests.

Our response rate is currently pretty good, but it’s unclear whether we can sustain it indefinitely. We all have day jobs, for example. One way you could help is to review the list of unchecked unblocks and gather up missing information, or look at the ones tagged moreinfo and see whether that’s still the case (maybe the maintainer replied, but forgot to remove the tag). If you’re confident, you might even try triaging some of the obvious requests and give some feedback to the maintainer, though the final decision will be made by a release team member.

After all, the quicker this goes the sooner we can release and thaw up unstable again.


Footnote: the method used to determine RC bug counts last week and this week differ, and therefore so could the margin for error. Surprisingly enough, counting bugs is not an exact science. I’m confident these numbers are close enough for broad comparison, even if they’re out by one or two.