Introducing: Bug horizon
I've come up with a new plan to organize potato's convergence toward zero
release-critical bugs, since the current approach doesn't seem to be
I'm going to take a snapshot of the release-critical bug list today.
Two weeks from now (February 28th), I will take another snapshot.
The 28th of February is the Bug Horizon.
After the Horizon, I will remove from potato every package that has
release-critical bugs that appear unchanged on both lists. I may make
exceptions for essential packages like libc6... perhaps :) Assume I'm
the Mad Release Manager, though.
This means that these two weeks are the time to fix things. If
you see a package on the list that you think should not be removed,
then jump in and do something about it!
A bugsquashing party will probably be scheduled during this time.
If you're working on a last-minute fix, or you have a strategy for
fixing a bug that will take a bit more time, then let me know or
send the info to email@example.com (where it will be added to
the bugscan report).
Two weeks from now, I expect some 150 new release-critical bugs, so I'll
probably set a new Bug Horizon after the removal.
Note: This approach ensures that packages will only be removed for
release-critical bugs that are at least two weeks old. I've noticed
that two weeks is a kind of cutoff point for bugs. Most of them get
fixed in two weeks; the ones that don't will take much longer.
In other news:
The response to my call for volunteers has been good. A "reminder
team" consisting of four people is forming:
I've had more volunteers after the initial rush, but I think this
team is about the right size now. (If not, I'll leave it up to
them to expand). If you want to help, there are lots of other
ways, now that reminder mails are being taken care of. Look at
the [HELP] tags in the bugscan report for example.
(Please remember that debian-devel-announce is an announcements list.
Discussions about announcements should be held on the appropriate list,