F13 Blocker bugs urgently need re-testing for tomorrow

[Peter Robinson]

On Tue, May 4, 2010 at 4:47 AM, Adam Williamson <awilliam at redhat.com> wrote:
> On Tue, 2010-05-04 at 00:45 +0100, Peter Robinson wrote:
>> On Tue, May 4, 2010 at 12:38 AM, Jesse Keating <jkeating at redhat.com> wrote:
>> > On Tue, 2010-05-04 at 00:03 +0100, Peter Robinson wrote:
>> >>
>> >> How are the pushes to updates-testing and stable working from within
>> >> bohdi for the release. I know we need to use it but compared to other
>> >> releases where the final packages have required rel-eng tickets I've
>> >> not seen anything other than "we're using bodhi".
>> >>
>> >> I've pushed some I need for the moblin spin (some were in testing) to
>> >> stable and they don't seem to have made it. Do they need blocker bugs
>> >> to be tagged into the final release now, rather than through bodhi?
>> >> Another bug for sugar that has a blocker bug has had the package
>> >> tagged through to final it seems.
>> >
>> > Until we enter RC phase, that is until the blocker list is cleared,
>> > we'll continue to use bodhi as the tool to move builds from candidate,
>> > to testing, to stable (/in/ the release).  Once we enter RC phase, a
>> > releng ticket or IRC bump will be needed to get something into the
>> > release.  However since once we hit RC phase, we're only taking things
>> > which are deemed to be release blocking issues, we should all be aware
>> > of the issue and be chomping at the bit for the fix.  Otherwise updates
>> > will be pushed via bodhi into the updates repo and be available as 0-day
>> > (and well -# day) updates.
>>
>> Excellent, thanks for the explanation.
>
> Just to add to what Jesse said, there've been some issues over the last
> few days, so you may have seen some push requests seemingly not working,
> or packages that are listed as having been 'pushed to stable' not being
> available from mirrors etc. As implied by what Jesse wrote, all of this
> behaviour so far is caused simply by problems, it's not an intentional
> change; regardless of the problems, the procedure is indeed just as it
> has been for months, you push to testing and then to stable after
> getting some feedback, just like with stable releases. There'll only be
> an intentional policy change - like Jesse describes - once we get an RC
> out (which is supposed to be tomorrow...but, well, look at the blocker
> list).
>
> Hopefully the problems with Koji/mash/whatever will be ironed out soon
> and all the packages which ought to be in dist-f13 will be there with
> tomorrow's/today's compose.

Hi Adam,

Thanks for the explanation, I had pretty much assumed the tech issues
as I'd seen things 'pushed' but then not updated etc. The reason I
queried was that I wasn't exactly sure when we would go from the 'push
comes from bodhi' to 'explicit tagging + day 0 updates from bodhi'.
I'm sure it had all been explained at some point but I suspect I
missed the email in the usual deluge :-)

Cheers,
Peter