On Thu, 2021-11-18 at 16:27 +0100, Fabio Valentini wrote: > On Thu, Nov 18, 2021 at 1:14 PM Sérgio Basto <sergio@xxxxxxxxxx> wrote: > > > > Hi, > > another subject that can be related > > > > In thread "I think we should stop building i686 packages we're not > > shipping" we are alerted for builds for i686 that aren't publish . > > Following the tip , I found some packages that we aren't shipped are > > needed to allow mock start the buildroot for example tar-1.34- > > 2.fc35.i686.rpm, tar is only available on koji local repo (1). > > > > I think we should have all i686 packages, that are needed to start a > > buildrooot, published. i.e. available on x86_64 repos. To allow mock > > build i686 packages without need to access to internal koji local > > repo > > (which may not be public) . > > This would break all sorts of things. > > It's also not compatible with the current packaging guidelines, > because packages must not conflict with each other unless they have an > explicit Conflicts tag and a good reason to do so. in this case, > tar.x86_64 and tar.i686 would conflict with each other (both providing > /usr/bin/tar, among other files), so that's out of the question. > > That's exactly the reason why complex heuristics are applied to which > i686 packages are allowed into the x86_64 repositories as "multilib" > packages. I think that is the way, only multilib package should build for i686 , and all packages that we need to start a build (buildroot) should be multilib . As is it done on RPMFusion, multilib packages have a special koji target that is set on rfpkg [1] and maybe this need to be set on a configuration file somewhere , I'd like that you give me suggestions, thank you. [1] https://github.com/rpmfusion-infra/rfpkg/blob/master/rfpkg/__init__.py#L194 > > But for packages that aren't shipped , I agree not build for i686, we > > already got the concept of multilib package and only multilib > > packages > > should be build for i686 , I think. > > Fabio > _______________________________________________ > devel mailing list -- devel@xxxxxxxxxxxxxxxxxxxxxxx > To unsubscribe send an email to devel-leave@xxxxxxxxxxxxxxxxxxxxxxx > Fedora Code of Conduct: > https://docs.fedoraproject.org/en-US/project/code-of-conduct/ > List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines > List Archives: > https://lists.fedoraproject.org/archives/list/devel@xxxxxxxxxxxxxxxxxxxxxxx > Do not reply to spam on the list, report it: > https://pagure.io/fedora-infrastructure -- Sérgio M. B. _______________________________________________ devel mailing list -- devel@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe send an email to devel-leave@xxxxxxxxxxxxxxxxxxxxxxx Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@xxxxxxxxxxxxxxxxxxxxxxx Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
