On Sat, Nov 02, 2013 at 08:02:51PM +0000, Richard Hughes wrote:
update fails, you either get corrupted data and crashing
or a hosed rpmdb. In a related point, we need to reduce the number of
updates we present to the user in a massive way in a supposedly
I think we probably _can_ find a good technical solution to the other. Or at
least, find some cases which we can handle reasonable well and work up fro
there. Running applications in containers will help immensely, but that's
kind of a long way off.
So, really, it's this related point that I'm concerned about now. We _need_
to do these things in coordination, not just push a situation into F20 where
we are telling our users to reboot everyday -- that's a pretty bad user
experience. Even weekly feels like a lot.
Matthew Miller ☁☁☁ Fedora Cloud Architect ☁☁☁ <mattdm(a)fedoraproject.org>