| *** brinzhang0 has quit IRC | 00:02 | |
| *** brinzhang has joined #openstack-oslo | 00:43 | |
| *** hamalq_ has quit IRC | 02:50 | |
| *** rcernin has quit IRC | 04:12 | |
| *** rcernin has joined #openstack-oslo | 04:18 | |
| *** larainema has joined #openstack-oslo | 04:19 | |
| *** zzzeek has quit IRC | 06:45 | |
| *** zzzeek has joined #openstack-oslo | 06:46 | |
| *** zzzeek has quit IRC | 07:51 | |
| *** zzzeek has joined #openstack-oslo | 07:53 | |
| *** hberaud has quit IRC | 07:59 | |
| *** hberaud has joined #openstack-oslo | 08:00 | |
| *** ralonsoh has joined #openstack-oslo | 08:02 | |
| *** rcernin has quit IRC | 08:07 | |
| *** sboyron has joined #openstack-oslo | 08:13 | |
| *** rpittau|afk is now known as rpittau | 08:30 | |
| *** sboyron has quit IRC | 08:38 | |
| *** rcernin has joined #openstack-oslo | 08:43 | |
| *** elod is now known as elod_pto | 09:04 | |
| *** rcernin_ has joined #openstack-oslo | 09:25 | |
| *** rcernin has quit IRC | 09:25 | |
| *** rcernin_ has quit IRC | 10:36 | |
| *** raildo has joined #openstack-oslo | 11:22 | |
| *** zzzeek has quit IRC | 11:28 | |
| *** rcernin_ has joined #openstack-oslo | 11:29 | |
| *** zzzeek has joined #openstack-oslo | 11:29 | |
| *** rcernin_ has quit IRC | 11:53 | |
| *** rcernin_ has joined #openstack-oslo | 11:53 | |
| *** brinzhang has quit IRC | 12:08 | |
| *** sboyron has joined #openstack-oslo | 12:10 | |
| ttx | Hi team, oslo-messaging (and I suspect other libs) are currently failing the lower-constraints job (even for changes that do not alter the lower-constraints), with: | 12:18 |
|---|---|---|
| ttx | ERROR: No matching distribution found for bandit<1.7.0,>=1.6.0 | 12:18 |
| ttx | It's a bit unclear to me why it would not find bandit 1.6.2 | 12:19 |
| ttx | Example fail at https://review.opendev.org/c/openstack/oslo.messaging/+/727787 | 12:20 |
| ttx | /cc fungi | 12:20 |
| ttx | (it's on master branch fwiw) | 12:22 |
| ttx | I can't reproduce the same error locally | 12:31 |
| *** sboyron has quit IRC | 12:33 | |
| hberaud | ttx: yep I started to submit patches to fix that unfortunatelly it's easy to fall in the dependency hell with this topic | 12:38 |
| hberaud | ttx: a lot of oslo project still need fix | 12:38 |
| hberaud | ttx: the method I used to reproduce that is 1. "tox -r -e venv -- pip install -U pip" 2. tox -e venv -- pip install -c lower-constraints.txt -r test-requirements.txt -r requirements.txt | 12:40 |
| hberaud | IIUC concerning bandit the issue is that bandit in lower-constraints and in test-requirements are not aligned | 12:41 |
| hberaud | or not enough aligned to allow the new pip resolver to resolve things correctly | 12:42 |
| hberaud | another big issue with oslo is the number of layers in requirements between oslo project, by example oslo.messaaaging depends on oslo.log wich depends on oslo.config etc... etc... so to fix one thing we need to start fixing at the right place first | 12:44 |
| *** zzzeek has quit IRC | 12:54 | |
| *** zzzeek has joined #openstack-oslo | 12:58 | |
| *** tosky has joined #openstack-oslo | 12:58 | |
| ttx | There were two suggestions in a recent thread: drop lower-constraints entirely (which some projects seem to have done), and move test-requirements to pre-venv tox.ini entries | 13:05 |
| ttx | per-venv* | 13:05 |
| ttx | Looks like the first one would "fix" this | 13:06 |
| ttx | I'm all for simplification... We have built something that is great on paper, but hard to keep in good shape and that only a limited number of people actually understand | 13:07 |
| ttx | Sure it was a great service as long as it lasted, but now it may be time to push that lower-bounds validation back in distroland. It's hard enough to keep up with the higher-bound as dependencies get updated | 13:09 |
| tosky | on the other hand, don't you think that once this big disruption coming from a one-time change of the pip resolver is fixed, the future maintenance will be easier? | 13:12 |
| hberaud | ttx: so it seems that the shortest way is to drop LC | 13:13 |
| ttx | tosky: imho lower-constraints is not facilitating anything but the case of people wanting to build a distro on old deps | 13:18 |
| ttx | Currently they can take our lower-constraints set and run with it. Tomorrow they will have to determine a set and run tests | 13:19 |
| ttx | I'm not sure keeping them around is making future maintenance of "openstack" easier. | 13:20 |
| ttx | Maintaining Lower constraints does not help is keeping track of new deps | 13:20 |
| hberaud | yes I don't expect that similar issue will happen so early, AFAIK the pip resolver topic is an old file... However I agree with ttx a simplification can't hurt. I saw divergence between upstream LC and version really available downstream, so it don't seems really respect outside upstream. | 13:20 |
| ttx | help in* | 13:20 |
| hberaud | s/respect/respected/ | 13:21 |
| *** zzzeek has quit IRC | 13:21 | |
| ttx | Distros have to test with their deps anyway | 13:21 |
| hberaud | so I'm not sure of benefit to continue with LC | 13:22 |
| ttx | But we'll likely need a TC decision to really drop them | 13:22 |
| hberaud | make sense | 13:23 |
| ttx | hberaud: I'd say there is still some fringe benefit (always was), but not enough to justify the complexity and recurring pain | 13:23 |
| hberaud | +1 | 13:23 |
| ttx | When we had dozens of people understanding and maintaining it, it was still a good tarde-off. Today not so much | 13:24 |
| hberaud | I see | 13:24 |
| *** zzzeek has joined #openstack-oslo | 13:25 | |
| *** zzzeek has quit IRC | 13:29 | |
| *** zzzeek has joined #openstack-oslo | 13:31 | |
| openstackgerrit | Hervé Beraud proposed openstack/openstack-doc-tools master: Dropping lower constraints testing https://review.opendev.org/c/openstack/openstack-doc-tools/+/767723 | 14:04 |
| openstackgerrit | Hervé Beraud proposed openstack/pbr master: Dropping lower constraints testing https://review.opendev.org/c/openstack/pbr/+/767724 | 14:04 |
| openstackgerrit | Hervé Beraud proposed openstack/automaton master: Dropping lower constraints testing https://review.opendev.org/c/openstack/automaton/+/767725 | 14:05 |
| openstackgerrit | Hervé Beraud proposed openstack/debtcollector master: Dropping lower constraints testing https://review.opendev.org/c/openstack/debtcollector/+/767727 | 14:06 |
| openstackgerrit | Hervé Beraud proposed openstack/futurist master: Dropping lower constraints testing https://review.opendev.org/c/openstack/futurist/+/767728 | 14:06 |
| openstackgerrit | Hervé Beraud proposed openstack/oslo.cache master: Dropping lower constraints testing https://review.opendev.org/c/openstack/oslo.cache/+/767729 | 14:07 |
| openstackgerrit | Hervé Beraud proposed openstack/oslo.concurrency master: Dropping lower constraints testing https://review.opendev.org/c/openstack/oslo.concurrency/+/767730 | 14:07 |
| openstackgerrit | Hervé Beraud proposed openstack/oslo.context master: Dropping lower constraints testing https://review.opendev.org/c/openstack/oslo.context/+/767731 | 14:07 |
| openstackgerrit | Hervé Beraud proposed openstack/oslo.db master: Dropping lower constraints testing https://review.opendev.org/c/openstack/oslo.db/+/767732 | 14:08 |
| openstackgerrit | Hervé Beraud proposed openstack/oslo.i18n master: Dropping lower constraints testing https://review.opendev.org/c/openstack/oslo.i18n/+/767733 | 14:08 |
| openstackgerrit | Hervé Beraud proposed openstack/oslo.limit master: Dropping lower constraints testing https://review.opendev.org/c/openstack/oslo.limit/+/767734 | 14:09 |
| openstackgerrit | Hervé Beraud proposed openstack/oslo.log master: Dropping lower constraints testing https://review.opendev.org/c/openstack/oslo.log/+/767735 | 14:09 |
| openstackgerrit | Hervé Beraud proposed openstack/oslo.messaging master: Dropping lower constraints testing https://review.opendev.org/c/openstack/oslo.messaging/+/767736 | 14:09 |
| openstackgerrit | Hervé Beraud proposed openstack/oslo.metrics master: Dropping lower constraints testing https://review.opendev.org/c/openstack/oslo.metrics/+/767737 | 14:10 |
| openstackgerrit | Hervé Beraud proposed openstack/oslo.middleware master: Dropping lower constraints testing https://review.opendev.org/c/openstack/oslo.middleware/+/767739 | 14:10 |
| openstackgerrit | Hervé Beraud proposed openstack/oslo.policy master: Dropping lower constraints testing https://review.opendev.org/c/openstack/oslo.policy/+/767740 | 14:10 |
| openstackgerrit | Hervé Beraud proposed openstack/oslo.privsep master: Dropping lower constraints testing https://review.opendev.org/c/openstack/oslo.privsep/+/767741 | 14:11 |
| openstackgerrit | Hervé Beraud proposed openstack/oslo.reports master: Dropping lower constraints testing https://review.opendev.org/c/openstack/oslo.reports/+/767742 | 14:11 |
| openstackgerrit | Hervé Beraud proposed openstack/oslo.rootwrap master: Dropping lower constraints testing https://review.opendev.org/c/openstack/oslo.rootwrap/+/767743 | 14:12 |
| openstackgerrit | Hervé Beraud proposed openstack/oslo.serialization master: Dropping lower constraints testing https://review.opendev.org/c/openstack/oslo.serialization/+/767744 | 14:12 |
| openstackgerrit | Hervé Beraud proposed openstack/oslo.service master: Dropping lower constraints testing https://review.opendev.org/c/openstack/oslo.service/+/767745 | 14:12 |
| openstackgerrit | Hervé Beraud proposed openstack/oslo.upgradecheck master: Dropping lower constraints testing https://review.opendev.org/c/openstack/oslo.upgradecheck/+/767746 | 14:13 |
| openstackgerrit | Hervé Beraud proposed openstack/oslo.utils master: Dropping lower constraints testing https://review.opendev.org/c/openstack/oslo.utils/+/767747 | 14:13 |
| openstackgerrit | Hervé Beraud proposed openstack/oslo.versionedobjects master: Dropping lower constraints testing https://review.opendev.org/c/openstack/oslo.versionedobjects/+/767748 | 14:14 |
| openstackgerrit | Hervé Beraud proposed openstack/oslo.vmware master: Dropping lower constraints testing https://review.opendev.org/c/openstack/oslo.vmware/+/767749 | 14:14 |
| openstackgerrit | Hervé Beraud proposed openstack/oslotest master: Dropping lower constraints testing https://review.opendev.org/c/openstack/oslotest/+/767750 | 14:14 |
| openstackgerrit | Hervé Beraud proposed openstack/osprofiler master: Dropping lower constraints testing https://review.opendev.org/c/openstack/osprofiler/+/767751 | 14:15 |
| openstackgerrit | Hervé Beraud proposed openstack/stevedore master: Dropping lower constraints testing https://review.opendev.org/c/openstack/stevedore/+/767752 | 14:15 |
| openstackgerrit | Hervé Beraud proposed openstack/taskflow master: Dropping lower constraints testing https://review.opendev.org/c/openstack/taskflow/+/767753 | 14:16 |
| hberaud | ttx: normally all oslo's LC tests are dropped by ^ (c.f => https://review.opendev.org/q/topic:"oslo_lc_drop") | 14:18 |
| ttx | oh, that was fast | 14:20 |
| ttx | I think we should wait for the TC to bless that dropping lower-constraints is a good idea | 14:20 |
| hberaud | sure np,,, patches are ready, we just have to push the right button (+W or abandon) | 14:21 |
| ttx | I might start making my patches depend on those ones though :P | 14:28 |
| hberaud | ttx: If that can reassure you, now I have to rebase all mine too :) | 14:32 |
| *** iurygregory has joined #openstack-oslo | 14:38 | |
| *** rcernin_ has quit IRC | 15:05 | |
| fungi | ttx: thanks for the example, and yeah the lower-constraints.txt files in most projects were not actually describing consistent satisfiable sets of dependency versions. i expected this was the case all along (and its part of why i repeatedly argued against adding lower-constraints jobs that way in the first place), but the new and more complete dep solver in pip 20.3 has made it clear to lots more | 15:07 |
| fungi | people that they have always been thoroughly broken | 15:07 |
| fungi | the jobs used to "succeed" but they didn't actually test with the versions of packages which were being requested | 15:09 |
| fungi | nor with the versions of packages other packages claimed were suitable for them | 15:09 |
| hberaud | yep | 15:11 |
| *** rcernin_ has joined #openstack-oslo | 15:22 | |
| *** rpittau is now known as rpittau|afk | 15:38 | |
| *** rcernin_ has quit IRC | 15:42 | |
| *** rcernin_ has joined #openstack-oslo | 15:44 | |
| bnemec | :-( | 16:04 |
| bnemec | lower-contraints might not have been doing exactly what it said, but it did catch a few lower bounds issues in my patches over the years. | 16:04 |
| bnemec | It's unlikely I'm going to fix the jobs though, so I guess that's the way it goes. | 16:05 |
| *** rcernin_ has quit IRC | 16:07 | |
| *** gibi is now known as gibi_pto | 16:35 | |
| hberaud | bnemec: things are not set in stone yet, there is still time to debate :) | 16:52 |
| gmann | hberaud: we can drop the lower-constraints.txt file and tox env also. | 17:12 |
| hberaud | gmann: ack | 17:13 |
| *** hamalq has joined #openstack-oslo | 17:19 | |
| *** hamalq_ has joined #openstack-oslo | 17:21 | |
| *** hamalq has quit IRC | 17:24 | |
| *** dave-mccowan has quit IRC | 17:42 | |
| *** dave-mccowan has joined #openstack-oslo | 17:47 | |
| *** ralonsoh has quit IRC | 17:57 | |
| openstackgerrit | Thierry Carrez proposed openstack/oslo.messaging master: Add Support For oslo.metrics https://review.opendev.org/c/openstack/oslo.messaging/+/761848 | 18:16 |
| *** dave-mccowan has quit IRC | 19:37 | |
| *** rcernin_ has joined #openstack-oslo | 20:42 | |
| *** raildo has quit IRC | 21:07 | |
| *** rcernin_ has quit IRC | 21:51 | |
| *** rcernin_ has joined #openstack-oslo | 22:29 | |
| *** rcernin_ has quit IRC | 22:34 | |
| *** rcernin has joined #openstack-oslo | 22:35 | |
| *** zzzeek has quit IRC | 23:27 | |
| *** zzzeek has joined #openstack-oslo | 23:28 | |
| *** zzzeek has quit IRC | 23:51 | |
| *** zzzeek has joined #openstack-oslo | 23:53 | |
| *** zzzeek has quit IRC | 23:58 | |
| *** zzzeek has joined #openstack-oslo | 23:59 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!