On Tue, Mar 07, 2006 at 04:10:29PM +0100, Marc Haber wrote: > On Wed, Mar 08, 2006 at 12:54:58AM +1000, Anthony Towns wrote: > > Dec 14th, 2.6.8 and 2.4.27 advisories get released, the first > > kernel updates for sarge > > Dec 17th, 3.1r1 gets released > > Dec 20th, 3.1r1 gets announced > > > > Jan 20th, DSA-946-1 is released for sudo, breaking the buildds, > > and introducing critical bugs 349196, 349549, 349587, 349729 > > Feb 6th, Joey mails indicating he'd like to release the update > > at the end of Feb (27th/28th) or a little bit later at > > the end of February. "let me know if this is ok for > > you - or if this is not ok for you" > > Feb 22nd, I mail both Joey (as SRM) and the security team noting the > > queue changes that should happen "with a stable update > > coming up" > I note that it took you 16 days to reply, It wasn't a reply to Joey's mail, which I hadn't thought had needed a reply. It was just a related mail about preparing the stable update. > and that you seem to want to > build a dependency between a change which is not strictly needed to > make a point release (if it were needed, why was it possible to > release 3.1r1?) and 3.1r2. May I ask why? The dependency is the other way -- that change needs to happen immediately after a stable update, and this is the first one that's suitable. The reason I think it's a good idea is that it means most of the work can be done by the stable release manager directly during the months in between updates, rather than as part of the update itself. Raphael's guess at why is correct -- so that proposed-updates really doesn't have anything the SRM would reject. Cheers, aj
Attachment:
signature.asc
Description: Digital signature