openstackgerrit | Merged openstack/requirements stable/train: Cap setuptools<46.0.0 https://review.opendev.org/712777 | 00:00 |
---|---|---|
*** tosky has quit IRC | 00:18 | |
*** vishalmanchanda has joined #openstack-requirements | 00:43 | |
openstackgerrit | Prashant Bhole proposed openstack/requirements master: Added sqlalchemy-filters requirement https://review.opendev.org/710555 | 01:01 |
rm_work | hey FYI eventlet==0.20.0 is broken for py36 which is used in a lot of projects in stable/rocky, I would like to propose bumping the upper-constraint to 0.21.0 in rocky, is that ... reasonable? | 03:55 |
prometheanfire | rm_work: isn't rocky close to dead? | 03:58 |
prometheanfire | ya | 03:59 |
prometheanfire | Extended Maintenance estimated 2020-02-24 | 03:59 |
prometheanfire | rm_work: I'd say probably not :| | 03:59 |
prometheanfire | that's a fairly large change for somethinglike eventlet | 03:59 |
prometheanfire | I'm happy to be told I'm being an idiot :P | 03:59 |
* prometheanfire wonders if the EM people should manage reqs for the EM branches | 03:59 | |
*** udesale has joined #openstack-requirements | 04:40 | |
rm_work | Nah you're probably right | 04:40 |
rm_work | Since that version bump changes a return type that needs an accompanying patch to at least one project, probably more | 04:41 |
rm_work | Guess we just have to patch downstream to make py36 work | 04:41 |
rm_work | Wish we could bump the minimum explicitly for a single project, but as it is, that doesn't work | 04:41 |
rm_work | Since updating L-C fails requirements check due to U-C in global | 04:42 |
*** bnemec has quit IRC | 05:19 | |
*** bnemec has joined #openstack-requirements | 05:28 | |
*** evrardjp has quit IRC | 05:35 | |
*** evrardjp has joined #openstack-requirements | 05:36 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/requirements master: Updated from generate-constraints https://review.opendev.org/712838 | 06:21 |
openstackgerrit | Merged openstack/requirements master: Updated from generate-constraints https://review.opendev.org/712615 | 07:20 |
*** rpittau|afk is now known as rpittau | 07:24 | |
*** e0ne has joined #openstack-requirements | 07:51 | |
*** e0ne has quit IRC | 07:53 | |
*** tosky has joined #openstack-requirements | 08:46 | |
*** e0ne has joined #openstack-requirements | 08:58 | |
*** ralonsoh has joined #openstack-requirements | 08:59 | |
*** udesale_ has joined #openstack-requirements | 11:09 | |
*** udesale has quit IRC | 11:12 | |
*** rpittau is now known as rpittau|bbl | 11:28 | |
*** udesale_ is now known as udesale | 12:16 | |
*** vishalmanchanda has quit IRC | 13:03 | |
*** rpittau|bbl is now known as rpittau | 13:04 | |
openstackgerrit | Sorin Sbarnea proposed openstack/requirements master: Fixed pytest-html constraints https://review.opendev.org/712942 | 13:08 |
zbr | when where the py27 jobs removed from reqs repo? bugs got in, https://review.opendev.org/#/c/712942/ | 13:11 |
zbr | i may be missing something but it makes no sense to have py27 specific conditions in constraints file but not to test them. | 13:12 |
*** panda has joined #openstack-requirements | 13:15 | |
*** weshay|ruck has joined #openstack-requirements | 13:44 | |
*** bnemec is now known as beekneemech | 14:10 | |
zbr | prometheanfire: ^ please, that his quite an issue. | 14:22 |
smcginnis | zbr: Why are there py2 jobs being run yet? | 14:58 |
*** opetrenko has joined #openstack-requirements | 15:03 | |
fungi | more discussion in #openstack-infra, but requirements changes still get cross-swift-py27 run on them | 15:22 |
fungi | requirements-tox-py27-check-uc was dropped mid-january with https://review.opendev.org/702478 | 15:22 |
fungi | when neutron-lib stopped working with 2.7 | 15:22 |
smcginnis | At some point in the near future, we do plan on removing those py2/py3 specifiers for anything not directly needed by swift. | 15:25 |
fungi | i expect reinstating requirements-tox-py27-check-uc would mean adding a ton of new python_version=='2.7' lines | 15:25 |
fungi | and yeah, at least as long as swift is running jobs on python 2.7 which use upper-constraints.txt we presumably can't remove every last one of those | 15:26 |
fungi | on the other hand, if swift is the only remaining project declaring it's supporting 2.7, maybe it should just import a 2.7-specific constraints list for its own jobs | 15:26 |
smcginnis | fungi: We may need to go that route. | 15:27 |
fungi | that might be simpler than trying to merge it with the one in the requirements repo | 15:27 |
fungi | the point of openstack/requirements upper-constraints.txt is to coordinate the same versions testing across multiple projects | 15:27 |
smcginnis | And presumably, the py2 upper constraints are fairly static at this point. | 15:28 |
fungi | and coordinating dependency versions across multiple projects is obviously not what's happening with the 2.7-specific entries if only swift uses them | 15:28 |
fungi | so it's basically a swift-specific list of exclusions being embedded in the requirements repo for no real benefit to anyone and only added compexity | 15:29 |
smcginnis | I think I'll propose a patch to see how that would look. | 15:29 |
*** ccamacho has quit IRC | 15:32 | |
openstackgerrit | Ivan Kolodyazhny proposed openstack/requirements master: Use pyScss2 and django-pyscss2 https://review.opendev.org/712976 | 15:35 |
smcginnis | https://review.opendev.org/#/c/712984/ | 15:45 |
prometheanfire | ohh, neato | 15:46 |
smcginnis | zbr: You may want to do something similar for whichever project you are still running py27 tests. ^ | 15:46 |
zbr | there are 15-20 repos at least, this would be a huge PITA | 15:50 |
zbr | and another question: why was tox-py27 removed without removing the py27 conditions from the reqs file? | 15:50 |
zbr | is like disabling the tests, which allows introduction of breaking changes. | 15:51 |
smcginnis | We want to have those breaking changes. | 15:51 |
prometheanfire | zbr: we are working on removing those as well | 15:51 |
smcginnis | https://governance.openstack.org/tc/goals/selected/ussuri/drop-py27.html | 15:51 |
smcginnis | We still need to clean some things up. | 15:51 |
smcginnis | But projects shouldn't be relying on global upper-constraints for py27 tests other than swift. | 15:52 |
smcginnis | Anyone else using it should have dropped py27 testing by now. | 15:52 |
openstackgerrit | Ivan Kolodyazhny proposed openstack/requirements master: Use pyScss2 and django-pyscss2 https://review.opendev.org/712976 | 16:06 |
fungi | zbr: to reiterate, tox-py27 was removed without removing the py27 conditions from the reqs file because swift was still using them (and the cross-swift-py27 job was providing coverage for the entries they're relying on) | 16:15 |
fungi | swift was the only project which responded to the (lengthy) discussions about dropping py27 support, so any other projects which didn't tell the tc they wanted to keep using py27 past ussuri milestone 1 are not known to anyone and therefore can't really complain | 16:17 |
*** openstackgerrit has quit IRC | 16:31 | |
*** rpittau is now known as rpittau|afk | 16:58 | |
*** opetrenko has quit IRC | 17:00 | |
*** e0ne has quit IRC | 17:31 | |
*** iurygregory has quit IRC | 17:32 | |
*** evrardjp has quit IRC | 17:35 | |
*** evrardjp has joined #openstack-requirements | 17:36 | |
*** udesale_ has joined #openstack-requirements | 17:52 | |
*** udesale has quit IRC | 17:54 | |
*** ralonsoh has quit IRC | 18:36 | |
*** udesale_ has quit IRC | 18:39 | |
*** e0ne has joined #openstack-requirements | 19:01 | |
*** openstackgerrit has joined #openstack-requirements | 19:07 | |
openstackgerrit | Ivan Kolodyazhny proposed openstack/requirements master: Use pyScss2 and django-pyscss2 https://review.opendev.org/712976 | 19:07 |
*** brinzhang_ has joined #openstack-requirements | 19:38 | |
*** brinzhang has quit IRC | 19:42 | |
*** e0ne has quit IRC | 21:46 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!