Jeff Spaleta wrote:
On Tue, Dec 9, 2008 at 10:11 PM, Les Mikesell <lesmikesell@xxxxxxxxx> wrote:
Known-bad in this scenario means detected after the initial rollout but
before the end of the window to the next 'new code' push. Being able to
avoid those and get the fixed version instead on your critical machines
would make using fedora a lot more practical.
Where do you collect and store the information that marks an update as
known bad?
Bugzilla? Where do you want people to report bugs? If you had a fixed
window before updates would be picked up on the machines with the more
conservative setting people might be more anxious to report them.
But I wouldn't envision marking an update as 'bad' although that's an
interesting concept itself. I was thinking that there would be a
specified time when all normal updates enter the repository, followed by
a time when only critical bug and security fix updates could be added,
so towards the end of that interval, packages that hadn't been replaced
with 'better' updates would automatically be assumed 'good' and it would
be fairly safe to update machines where you want less risk. Then a new
cycle of 'new feature' updates could start.
--
Les Mikesell
lesmikesell@xxxxxxxxx
--
fedora-devel-list mailing list
fedora-devel-list@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/fedora-devel-list
