I'm releasing a new version of the spin-kickstarts package today.
Please verify it's contents using the following build:
http://koji.fedoraproject.org/koji/taskinfo?taskID=1357916
Please ACK the change in the ticket I've logged with rel-eng to tag for dist-f11:
https://fedorahosted.org/rel-eng/ticket/1826
Kind regards,
Jeroen van Meeuwen -kanarip
On Sat, May 16, 2009 at 14:19:14 +0200, Jeroen van Meeuwen kanarip@kanarip.com wrote:
I'm releasing a new version of the spin-kickstarts package today.
Please verify it's contents using the following build:
Depending on far people are going to go with this, I'll note there is an issue with pykickstart and / or livecd-tools. The most recent update of pykickstart is broken or changes an interface used by livecd-tools that results in the --size parameter to part commands being ignored. You get the default of 4096, which isn't big enough for at least some DVD spins.
On 05/16/2009 07:10 PM, Bruno Wolff III wrote:
On Sat, May 16, 2009 at 14:19:14 +0200, Jeroen van Meeuwenkanarip@kanarip.com wrote:
I'm releasing a new version of the spin-kickstarts package today.
Please verify it's contents using the following build:
Depending on far people are going to go with this, I'll note there is an issue with pykickstart and / or livecd-tools. The most recent update of pykickstart is broken or changes an interface used by livecd-tools that results in the --size parameter to part commands being ignored. You get the default of 4096, which isn't big enough for at least some DVD spins.
Is this bug referenced somewhere? Are we / some of us tracking it?
Kind regards,
Jeroen van Meeuwen -kanarip
On Mon, May 18, 2009 at 18:37:38 +0200, Jeroen van Meeuwen kanarip@kanarip.com wrote:
Is this bug referenced somewhere? Are we / some of us tracking it?
https://bugzilla.redhat.com/show_bug.cgi?id=501020
It's filed under pykickstart, but it may really be an interface change, rather than a bug, that requires an update of livecd-tools.
On 05/18/2009 09:48 PM, Bruno Wolff III wrote:
On Mon, May 18, 2009 at 18:37:38 +0200, Jeroen van Meeuwenkanarip@kanarip.com wrote:
Is this bug referenced somewhere? Are we / some of us tracking it?
https://bugzilla.redhat.com/show_bug.cgi?id=501020
It's filed under pykickstart, but it may really be an interface change, rather than a bug, that requires an update of livecd-tools.
I think right now this has been fixed, built and pushed out to dist-f11-final, am I not mistaken?
Kind regards,
Jeroen van Meeuwen -kanarip
On Wed, May 20, 2009 at 16:46:12 +0200, Jeroen van Meeuwen kanarip@kanarip.com wrote:
On 05/18/2009 09:48 PM, Bruno Wolff III wrote:
On Mon, May 18, 2009 at 18:37:38 +0200, Jeroen van Meeuwenkanarip@kanarip.com wrote:
Is this bug referenced somewhere? Are we / some of us tracking it?
https://bugzilla.redhat.com/show_bug.cgi?id=501020
It's filed under pykickstart, but it may really be an interface change, rather than a bug, that requires an update of livecd-tools.
I think right now this has been fixed, built and pushed out to dist-f11-final, am I not mistaken?
Yes. But there is also a new problem with e2fsprogs. There is a test fix available in the bug (501198) that seems like it might solve that issue. (I just finished a successful test, but want to try a couple of more.)
There is also an error reported by rpm at the end, but I don't think that is messing up the iso. (But I haven't done a real test since that error started showing up.)
On Wed, 20 May 2009 10:12:55 -0500, Bruno Wolff III bruno@wolff.to wrote:
On Wed, May 20, 2009 at 16:46:12 +0200, Jeroen van Meeuwen kanarip@kanarip.com wrote:
On 05/18/2009 09:48 PM, Bruno Wolff III wrote:
On Mon, May 18, 2009 at 18:37:38 +0200, Jeroen van Meeuwenkanarip@kanarip.com wrote:
Is this bug referenced somewhere? Are we / some of us tracking it?
https://bugzilla.redhat.com/show_bug.cgi?id=501020
It's filed under pykickstart, but it may really be an interface change, rather than a bug, that requires an update of livecd-tools.
I think right now this has been fixed, built and pushed out to dist-f11-final, am I not mistaken?
Yes. But there is also a new problem with e2fsprogs. There is a test fix available in the bug (501198) that seems like it might solve that issue. (I just finished a successful test, but want to try a couple of more.)
There is also an error reported by rpm at the end, but I don't think that is messing up the iso. (But I haven't done a real test since that error started showing up.)
On behalf of all spin maintainers and the SIG;
Thanks for being on top of these issues Bruno!
--Jeroen
On Wed, 2009-05-20 at 17:23 +0200, Jeroen van Meeuwen wrote:
On behalf of all spin maintainers and the SIG;
Thanks for being on top of these issues Bruno!
Yes, thank you Bruno, you've saved me a lot of heartache.
On Wed, May 20, 2009 at 09:45:14 -0700, Jesse Keating jkeating@redhat.com wrote:
On Wed, 2009-05-20 at 17:23 +0200, Jeroen van Meeuwen wrote:
On behalf of all spin maintainers and the SIG;
Thanks for being on top of these issues Bruno!
Yes, thank you Bruno, you've saved me a lot of heartache.
It's mostly been selfish. I keep trying to get Fedora install at my Thursday night computer gaming group while we aren't actually using the machines to play games and things have broken the last few weeks and I didn't have enough time to get things working in a small amount of time. And then the next week I would want to have the latest updates, so I would make new spins and find a different issue. It's been frustrating, but I seem to have a masochistic streak when it comes to computers.
P.S. I already sent a correction to Jeroen; the bug for e2fsprogs issues (which are not completely fixed as my last test failed) is 499452 not 501198. (501198 is a mkinitrd bug affecting some systems with encrypted partitions and shouldn't be a problem for live spins, but did give me some grief.) Also there is a bug introduced by a recent yum update, that I don't think is blocking builds but might be interacting with the e2fsprogs issue. This one is 501606 and has a suggested change, but no actual update yet.
-
Bruno Wolff III -
Jeroen van Meeuwen -
Jesse Keating