[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Debian Re-organization proposals



In article <87vhqlej5q.fsf@tiamat.datasync.com> you wrote:

: 	This is predicated on the assumption that maintainers are
:  going to exercise judgement in promoting a package to stable; things
:  break down (but no worse than the current state) if people, either in
:  error, or impatience, promote packages to stable.

It is conceivable that moving a package to a declared "stable" state could
require that the package revision not have been superceded for some period
of time, that there be no new bugs filed against that version of the package,
and that the maintainer says "ok".  Exceptions, as always, possible but
frowned upon.

: 	I vote we keep the previous stable version of a package in
:  safe storage until the newly promoted package has had time to mature
:  and perhaps be tested.

Yes.  A time-limited repository of the package revs that have been replaced
in the stable tree probably makes lots of sense.

: 	There are problems with this approach -- what about shared
:  libraries? 

Easy.  Force pkg-order congruence before something hits stable.  In other
words, a package can't go stable unless all of its dependencies have also 
gone stable.  Then, just don't sweat the small stuff... if there's some funny
change between a frozen and unstable library that tweaks a bug in a package,
we go "oops", fix it, and remind people that in this new world order, 
"stable" != "tested rigorously".

Bdale


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org


Reply to: