On Tue, Apr 06, 2004 at 08:20:54AM -0400, David B Harris wrote: > On Tue, 6 Apr 2004 01:15:13 -0500 > Steve Langasek <vorlon@debian.org> wrote: > > > Can one of the release managers comment on this? Me and the other ALSA > > > maintainers had planned on changing (perhaps getting rid of) the init > > > script for Sarge+1. If we need to accelerate these plans, we wouldn't > > > mind a bit of notice. It's a fairly major change. > > There seems to have been quite a bit of activity on this bug since your > > message; are you still looking for comments from the release team? > The hotplug team really wants the new package to get into testing, so > we've decided to go ahead and make our changes - they're still sort of > major, though. If you guys want to comment on whether that's a good or > bad idea at this point, feel free :) Go wild -- just don't introduce any new bugs! ;) > > I would note that any bug that can be described as "fake bug to keep foo > > out of testing until bar is ready" is almost certain to cause problems > > with partial upgrades. Although generally discouraged by policy, would > > a versioned Conflicts be appropriate here? (I.e., do you regard the > > current versions of hotplug and alsa-base to be usable enough that users > > should be able to install them together? Bug #241225 suggests the > > answer is no.) > Yeah, there will be a conflicts as well, but nobody (on either the > hotplug team or the ALSA team) wants to have to tell users "well, you > can either use hotplug or you can use ALSA", hence the fake bug to keep > hotplug out of testing until ALSA gets changed. The normal effect of such a conflict is to impose the restriction "you can't upgrade hotplug without removing alsa", which I trust most users of testing can cope with reasonably when it scrolls across their screen. Want to keep alsa? Ok, just don't upgrade hotplug for the time being. (IIRC, apt-get upgrade will naturally DTRT here.) -- Steve Langasek postmodern programmer
Attachment:
signature.asc
Description: Digital signature