Thursday, 2026-06-25

opendevreviewOpenStack Proposal Bot proposed openstack/requirements stable/2026.1: update constraint for os-ken to new release 4.1.2  https://review.opendev.org/c/openstack/requirements/+/99485708:05
opendevreviewOpenStack Proposal Bot proposed openstack/requirements master: update constraint for networking-generic-switch to new release 10.0.0  https://review.opendev.org/c/openstack/requirements/+/99486509:24
*** ykarel__ is now known as ykarel09:51
opendevreviewMerged openstack/requirements master: Updated from generate-constraints  https://review.opendev.org/c/openstack/requirements/+/99416911:02
opendevreviewMerged openstack/requirements stable/2026.1: update constraint for os-ken to new release 4.1.2  https://review.opendev.org/c/openstack/requirements/+/99485711:02
fricklerfyi I've proposed to switch the reqs update cron to monthly in an attempt to reduce reviewer load (mostly my own ;) https://review.opendev.org/c/openstack/project-config/+/99492314:54
stephenfinfrickler: out of curiosity, what are the manual updates that are currently needed?15:08
stephenfin(I have seen your manual updates to the patches. I'm just missing the context on why they're needed)15:09
fricklerstephenfin: I try to list the reasons in the commit message. for pyroute2 there is a bug in os-vif/nova/neutron that prevents using a newer version. maybe that cap could be hardcoded into one of the projects. then latest pytest causes swift unit tests to fail, so there is a manual cap to allow at least the remainder of the bumps to merge15:32
fricklerthe alternatives would be to not merge any updates until all issues are fixed, which would mean we could maybe have updates once per year. or have the bot propose one patch per library bumped, but that would end up in review- and merge-conflict hell15:33
stephenfinfrickler: would I be stepping on toes if I took a look at automating those downgrades tomorrow?16:19
stephenfinI assume you'd be okay keeping it weekly if we could make the process less manual?16:20
fricklerwell most of the time the work is spent finding out what to upgrade/pin, not then performing the actual pin. but I'm open to look at what you come up with16:43
mnasiadkaOne patch per library wouldn’t be bad, but merge-conflict hell is surely not what we want to manage here17:03
fricklerthe insane part about that is, in reality there are no merge-conflicts for upper-constraints.txt, the relevant context is 0, which can be told to e.g. git-rebase as "-C0", but good luck telling gerrit and zuul about that18:33

Generated by irclog2html.py 4.1.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!