[Date Prev][Date Next] [Thread Prev][Thread Next]
[Thread Index]
[Date Index]
[Author Index]
Re: [PATCH] Do not show disabled repos.
- From: Chuck Anderson <cra WPI EDU>
- To: anaconda-devel-list redhat com
- Subject: Re: [PATCH] Do not show disabled repos.
- Date: Wed, 19 Nov 2008 00:48:23 -0500
On Tue, Nov 18, 2008 at 07:45:49PM -1000, David Cantrell wrote:
> On Tue, Nov 18, 2008 at 09:21:51PM -0800, Jesse Keating wrote:
> > On Wed, 2008-11-19 at 00:17 -0500, Chris Lumens wrote:
> > >
> > > Also really the only time we can ever get testing on this is if it's
> > > enabled on a release. If we always say "oh it still doesn't work,
> > > disable it for release" then we'll never actually get it into a working
> > > state.
> >
> > You can get testing on this at any time you want. All it takes is a
> > special fedora-release and a special pungi call. These things take all
> > of about an hour to do. Or it you could have tested any one of the
> > alpha/beta/snapshot/prereleases that had media since they all would have
> > the rawhide box to check, which would emulate having the updates box at
> > release time.
> >
> > Claiming that you could only test at release time doesn't work here.
>
> Rather than pointing fingers at each other, we should probably work on adding
> this particular situation in to the F-11 test plan. I'm sure the QA guys are
> looking for some more work. :)
>
> > Of course, my suggestion of throwing it out isn't exactly helpful. My
> > patch tries to minimize the possibility of users shooting themselves in
> > the foot for the F10 release, so that something better can be achieved
> > in the F11 release. It's really too bad we have so little options at
> > this point, but it is what it is.
>
> I understand what you're trying to do, but here's my mindset at this point for
> F-10 anaconda:
>
> - Does the issue prevent people from installing?
> - If yes, make a patch, test the hell out it, have the team review it,
> get it in.
> - If no, it's too late for F-10.
>
> Really, I just hate the last minute changes that we've done to anaconda in the
> past. There's almost always something that slips in at the last minute and a
> new failure shows up and we're playing the holy-crap-fix-it game at the very
> end of the release cycle. I just want to avoid that entirely by sticking with
> the blocker/non-blocker concept, so I just consider this one non-blocker.
> Installs still proceed.
How about creating a new fedora-rawhide subpackage of fedora-release
to hold the rawhide repo files and don't install that by default?
[Date Prev][Date Next] [Thread Prev][Thread Next]
[Thread Index]
[Date Index]
[Author Index]