3:53 p.m.
On Wed, 2013-06-26 at 15:36 -0500, Bruno Wolff III wrote:
On Wed, Jun 26, 2013 at 13:30:50 -0700,
Adam Williamson <awilliam(a)redhat.com> wrote:
>In case you're asking the question in the topic: you may notice there is
>now an 'RC2.1' in the TC/RC download directory.
I was wondering about this.
You don't happen to know if a later version of spin-kickstarts was used
for this? (I suspect that is likely since there was a commit made to
fix one of the spins commited recently.) If so what do we want to do
for the spin-kickstarts package?
Well. It looks like Dennis used a later checkout. See:
http://koji.fedoraproject.org/koji/taskinfo?taskID=5547289 - that's the
RC2.1 x86_64 KDE compose. Note:
fedora-livecd-kde-96833b7.ks
that means it was built with spin-kickstarts git
96833b75f97d270a6057771f1b4a9da636350a83.
The KDE live kickstart did not change between 15305b2 (the commit we
used for RC2) and 96833b7 , so on that front we're fine, but the
fedora-live-jam-kde.ks kickstart *did* change: it got:
commit 834e2937b4294b5557b3e67da7ce219d152804e5
Author: Brendan Jones <brendan.jones.it(a)gmail.com>
Date: Wed Jun 26 07:26:12 2013 +0200
Jam: Remove slv2 and dssi-vst
So if we push the current spin-kickstarts to stable we are lying in an
extremely minor way about the final release (assuming for the moment we
go ahead and release RC2+RC2.1): the spin-kickstarts package will not
contain the actual fedora-live-jam.ks package used to build the RC2.1
live-jam spin.
There's two ways we could address this. We could build a new
spin-kickstarts with the changed fedora-live-jam.ks file *but no other
changes post-15305b2* (as then we'd be out of line on the other images).
Or we could just ship the RC2 version of the live-jam spin instead of
the RC2.1 version. I don't see that that would be a huge problem.
--
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | identi.ca: adamwfedora
http://www.happyassassin.net
4:02 p.m.
New subject: [Test-Announce] Fedora 19 Final RC2.1 WTF is this?!
On Wed, Jun 26, 2013 at 13:53:58 -0700,
Adam Williamson <awilliam(a)redhat.com> wrote:
There's two ways we could address this. We could build a new
spin-kickstarts with the changed fedora-live-jam.ks file *but no other
changes post-15305b2* (as then we'd be out of line on the other images).
Or we could just ship the RC2 version of the live-jam spin instead of
the RC2.1 version. I don't see that that would be a huge problem.
I'd rather have the release have the RC2 version. I can do a 0 day update
to get the fix for the jam spin.
4:09 p.m.
New subject: [Test-Announce] Fedora 19 Final RC2.1 WTF is this?!
On Wed, 2013-06-26 at 16:02 -0500, Bruno Wolff III wrote:
On Wed, Jun 26, 2013 at 13:53:58 -0700,
Adam Williamson <awilliam(a)redhat.com> wrote:
>
>There's two ways we could address this. We could build a new
>spin-kickstarts with the changed fedora-live-jam.ks file *but no other
>changes post-15305b2* (as then we'd be out of line on the other images).
>Or we could just ship the RC2 version of the live-jam spin instead of
>the RC2.1 version. I don't see that that would be a huge problem.
I'd rather have the release have the RC2 version. I can do a 0 day update
to get the fix for the jam spin.
Actually, hold that thought - RC2.1 turns out to be busted anyway. It
does not in fact include the updated liveusb-creator. They managed to
leave it out. =)
I'll get it redone (RC2.2?) and have it done with the 15305b2 kickstarts
this time, so that lines up.
--
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | identi.ca: adamwfedora
http://www.happyassassin.net
4:19 p.m.
New subject: [Test-Announce] Fedora 19 Final RC2.1 WTF is this?!
On Wed, 2013-06-26 at 14:09 -0700, Adam Williamson wrote:
On Wed, 2013-06-26 at 16:02 -0500, Bruno Wolff III wrote:
> On Wed, Jun 26, 2013 at 13:53:58 -0700,
> Adam Williamson <awilliam(a)redhat.com> wrote:
> >
> >There's two ways we could address this. We could build a new
> >spin-kickstarts with the changed fedora-live-jam.ks file *but no other
> >changes post-15305b2* (as then we'd be out of line on the other images).
> >Or we could just ship the RC2 version of the live-jam spin instead of
> >the RC2.1 version. I don't see that that would be a huge problem.
>
> I'd rather have the release have the RC2 version. I can do a 0 day update
> to get the fix for the jam spin.
Actually, hold that thought - RC2.1 turns out to be busted anyway. It
does not in fact include the updated liveusb-creator. They managed to
leave it out. =)
I'll get it redone (RC2.2?) and have it done with the 15305b2 kickstarts
this time, so that lines up.
Dennis is going to do the new build (RC2.2 or whatever) with 15305b2.
That'll mean the kde-jam spin doesn't get built, but it means we don't
have a problem with the spin-kickstarts package.
--
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | identi.ca: adamwfedora
http://www.happyassassin.net
11:22 a.m.
New subject: [Test-Announce] Fedora 19 Final RC2.1 WTF is this?!
On Wed, Jun 26, 2013 at 14:19:15 -0700,
Adam Williamson <awilliam(a)redhat.com> wrote:
Dennis is going to do the new build (RC2.2 or whatever) with 15305b2.
That'll mean the kde-jam spin doesn't get built, but it means we don't
have a problem with the spin-kickstarts package.
OK. I won't do any new spin-kickstart builds at this time.
7:14 p.m.
New subject: [Test-Announce] Fedora 19 Final RC2.1 WTF is this?!
On Thu, 2013-06-27 at 11:22 -0500, Bruno Wolff III wrote:
On Wed, Jun 26, 2013 at 14:19:15 -0700,
Adam Williamson <awilliam(a)redhat.com> wrote:
>
>Dennis is going to do the new build (RC2.2 or whatever) with 15305b2.
>That'll mean the kde-jam spin doesn't get built, but it means we don't
>have a problem with the spin-kickstarts package.
OK. I won't do any new spin-kickstart builds at this time.
Unfortunately, it seems you'll need to. 0.19.7-1 is bad, it does not
contain the right files. Details in my Bodhi comment:
https://admin.fedoraproject.org/updates/FEDORA-2013-11738/spin-kickstarts... .
--
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | identi.ca: adamwfedora
http://www.happyassassin.net
11:15 p.m.
New subject: [Test-Announce] Fedora 19 Final RC2.1 WTF is this?!
On Thu, Jun 27, 2013 at 17:14:14 -0700,
Adam Williamson <awilliam(a)redhat.com> wrote:
On Thu, 2013-06-27 at 11:22 -0500, Bruno Wolff III wrote:
> On Wed, Jun 26, 2013 at 14:19:15 -0700,
> Adam Williamson <awilliam(a)redhat.com> wrote:
> >
> >Dennis is going to do the new build (RC2.2 or whatever) with 15305b2.
> >That'll mean the kde-jam spin doesn't get built, but it means we
don't
> >have a problem with the spin-kickstarts package.
>
> OK. I won't do any new spin-kickstart builds at this time.
Unfortunately, it seems you'll need to. 0.19.7-1 is bad, it does not
contain the right files. Details in my Bodhi comment:
https://admin.fedoraproject.org/updates/FEDORA-2013-11738/spin-kickstarts... .
The way I did this was to use git reset to get to the desired commit and
then I had to update stuff and do more commits, since configure.ac and
spin-kickstarts.spec.in changes should also be in the build. Then I built
the tar ball which was used in the package update. I'll double check that
this is what actually happened.
11:20 p.m.
New subject: [Test-Announce] Fedora 19 Final RC2.1 WTF is this?!
On Thu, Jun 27, 2013 at 17:14:14 -0700,
Adam Williamson <awilliam(a)redhat.com> wrote:
On Thu, 2013-06-27 at 11:22 -0500, Bruno Wolff III wrote:
> On Wed, Jun 26, 2013 at 14:19:15 -0700,
> Adam Williamson <awilliam(a)redhat.com> wrote:
> >
> >Dennis is going to do the new build (RC2.2 or whatever) with 15305b2.
> >That'll mean the kde-jam spin doesn't get built, but it means we
don't
> >have a problem with the spin-kickstarts package.
>
> OK. I won't do any new spin-kickstart builds at this time.
Unfortunately, it seems you'll need to. 0.19.7-1 is bad, it does not
contain the right files. Details in my Bodhi comment:
https://admin.fedoraproject.org/updates/FEDORA-2013-11738/spin-kickstarts... .
Maybe I used the wrong options for git reset and the working files didn't
get changed or something. But things are defintely wrong. I'll do another
version.
11:50 p.m.
New subject: [Test-Announce] Fedora 19 Final RC2.1 WTF is this?!
On Thu, Jun 27, 2013 at 17:14:14 -0700,
Adam Williamson <awilliam(a)redhat.com> wrote:
On Thu, 2013-06-27 at 11:22 -0500, Bruno Wolff III wrote:
> On Wed, Jun 26, 2013 at 14:19:15 -0700,
> Adam Williamson <awilliam(a)redhat.com> wrote:
> >
> >Dennis is going to do the new build (RC2.2 or whatever) with 15305b2.
> >That'll mean the kde-jam spin doesn't get built, but it means we
don't
> >have a problem with the spin-kickstarts package.
>
> OK. I won't do any new spin-kickstart builds at this time.
Unfortunately, it seems you'll need to. 0.19.7-1 is bad, it does not
contain the right files. Details in my Bodhi comment:
https://admin.fedoraproject.org/updates/FEDORA-2013-11738/spin-kickstarts... .
I think I got it correct this time.
Hopefully by f20 final freeze I'll have the spin-kickstart build process
not needing to make commits as part of the build process. This will make
it easier to handle cases where people make commits after images are
built but before the matching spin-kickstarts package is built.
1:02 a.m.
New subject: [Test-Announce] Fedora 19 Final RC2.1 WTF is this?!
On Thu, 2013-06-27 at 23:50 -0500, Bruno Wolff III wrote:
On Thu, Jun 27, 2013 at 17:14:14 -0700,
Adam Williamson <awilliam(a)redhat.com> wrote:
>On Thu, 2013-06-27 at 11:22 -0500, Bruno Wolff III wrote:
>> On Wed, Jun 26, 2013 at 14:19:15 -0700,
>> Adam Williamson <awilliam(a)redhat.com> wrote:
>> >
>> >Dennis is going to do the new build (RC2.2 or whatever) with 15305b2.
>> >That'll mean the kde-jam spin doesn't get built, but it means we
don't
>> >have a problem with the spin-kickstarts package.
>>
>> OK. I won't do any new spin-kickstart builds at this time.
>
>Unfortunately, it seems you'll need to. 0.19.7-1 is bad, it does not
>contain the right files. Details in my Bodhi comment:
>https://admin.fedoraproject.org/updates/FEDORA-2013-11738/spin-kickstarts-0.19.7-1.fc19
.
I think I got it correct this time.
Hopefully by f20 final freeze I'll have the spin-kickstart build process
not needing to make commits as part of the build process. This will make
it easier to handle cases where people make commits after images are
built but before the matching spin-kickstarts package is built.
Yeah, splitting that out would help. We were kicking around other ideas
for improving the process in -releng today, we should try and thrash
that out before F20 Alpha.
--
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | identi.ca: adamwfedora
http://www.happyassassin.net
9:16 a.m.
New subject: [Test-Announce] Fedora 19 Final RC2.1 WTF is this?!
On Thu, Jun 27, 2013 at 23:02:44 -0700,
Adam Williamson <awilliam(a)redhat.com> wrote:
On Thu, 2013-06-27 at 23:50 -0500, Bruno Wolff III wrote:
Yeah, splitting that out would help. We were kicking around other ideas
for improving the process in -releng today, we should try and thrash
that out before F20 Alpha.
I started a thread (RFC 4) about it on the spins list.
I should have RFCs 1, 2, and 3 done by f20 alpha. I didn't want to start
mucking with the master branch too much, before f19 gold since it would
make it harder to make comparable changes to both f19 and master and
people were making a lot of changes down the home stretch.
4:24 p.m.
New subject: [Test-Announce] Fedora 19 Final RC2.1 WTF is this?!
On Thu, Jun 27, 2013 at 23:02:44 -0700,
Adam Williamson <awilliam(a)redhat.com> wrote:
Yeah, splitting that out would help. We were kicking around other ideas
for improving the process in -releng today, we should try and thrash
that out before F20 Alpha.
So, I ended up working on spin-kickstarts today and changed it so that
we don't need to make commits to make a new build. A new tag should be
pushed and the generated tar ball should get put on fedorahosted, but
neither of these affect the commit itself. The tar ball is then used
for the package update.
That should make doing these updates a lot simpler and less prone to
mistakes.
3961
days inactive
3964
days old
11 comments
2 participants
participants (2)
-
Adam Williamson -
Bruno Wolff III