On Mon, 11 Mar 2019 at 10:54, Troy Dawson <tdawson(a)redhat.com> wrote:
On Sat, Mar 9, 2019 at 8:16 PM Kevin Fenzi <kevin(a)scrye.com> wrote:
>
> On 3/9/19 7:15 PM, Troy Dawson wrote:
> > Hello,
> > There are a few questions I have, and since I'm not positive who all
> > of the correct people ask are, I'm sending to the epel-devel list.
> >
> > Before I start the questions, thank you to everyone who helped out in
> > getting the build failures working. Thank you. thank you. thank you.
> >
> > We have all the python packages rebuilt. All of the packages needed
> > for the rebuilds are tagged into override, and thus in the buildroot.
> > None of these rebuilt packages have been put into updates, and thus
> > are not in -testing.
> >
> > Where do we go from here?
> > A) Do I do one big update for all of the rebuilt packages?
>
> Yes.
>
> > B) Do I do individual updates for each package?
> > C) Some combination of above?
>
> Bad idea, as then some dependent packages could go out stable without
> all the packages they need.
>
> > D) Do a side tag, and take out all the overrides?
>
> Nope... side tag won't push them out, unless we just merge them into the
> main tag, but then they are bypassing testing.
>
> > I personally think C. Do the main components (rpm macros, python34
> > and python36) in one update. All the rest get their individual
> > update.
>
> All the other ones are going to need at least python36, so no, they
> should all be in one big update. Otherwise some of them could go stable
> and not have the python36 yet.
>
> > How long do we expect to be in this testing stage?
>
> I'd say at least 2 weeks?
>
> > If packagers want to update their python package for whatever reason,
> > what should they do?
>
> Build as usual, but ask someone to edit the update and remove the
> previous one and add in the new one.
>
> Thanks for building and coordinating everything!
>
> kevin
>
Thanks for the feedback. You too Miro.
After reading both of your reasons for doing A), I agree with you both.
Unless anyone objects, we'll plan to do option A
We only have one package left that fails it's rebuild, python-apsw.
I'm hoping someone can figure out the reason for the %checks failling. [1]
If we can get that fixed, I'll be all set to do the mass update later today.
If we are only going to have these builds in -testing for a couple of
weeks, how about we shoot for April 2.
I'll be unavailable from March 20-April 2. That would give us 3 weeks.
Hopefully by then we would have announcements made letting people know
the change is coming.
Sounds good. The possible 'fix' was to push to a newer version and
then edit it to not look at the older version of sqlite.
https://src.fedoraproject.org/fork/kanarip/rpms/
python-apsw/c/3fdeffe3e799e2305c821eb7d1bb710531ef7efe?branch=epel7-py36-by-apsw
38
Troy
[1] -
https://koji.fedoraproject.org/koji/taskinfo?taskID=33397847
_______________________________________________
epel-devel mailing list -- epel-devel(a)lists.fedoraproject.org
To unsubscribe send an email to epel-devel-leave(a)lists.fedoraproject.org
Fedora Code of Conduct:
https://getfedora.org/code-of-conduct.html
List Guidelines:
https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives:
https://lists.fedoraproject.org/archives/list/epel-devel@lists.fedoraproj...
--
Stephen J Smoogen.