| *** sboyron has quit IRC | 00:07 | |
| *** weshay|ruck is now known as weshay|pto | 00:41 | |
| *** dklyle has quit IRC | 00:48 | |
| *** ociuhandu has joined #openstack-infra | 01:03 | |
| *** ociuhandu has quit IRC | 01:08 | |
| *** vesper11 has joined #openstack-infra | 01:09 | |
| *** Goneri has quit IRC | 01:27 | |
| *** rcernin has quit IRC | 01:37 | |
| *** rcernin has joined #openstack-infra | 01:39 | |
| *** bdodd has quit IRC | 01:57 | |
| *** carloss has quit IRC | 02:02 | |
| *** rlandy has quit IRC | 02:15 | |
| *** ysandeep|away is now known as ysandeep | 02:18 | |
| *** ajitha has joined #openstack-infra | 02:32 | |
| *** rcernin has quit IRC | 02:36 | |
| *** rfolcOUT has joined #openstack-infra | 02:50 | |
| *** ociuhandu has joined #openstack-infra | 02:52 | |
| *** ociuhandu has quit IRC | 02:56 | |
| *** iurygregory|pto has quit IRC | 03:01 | |
| *** rfolcOUT has quit IRC | 03:05 | |
| *** rcernin has joined #openstack-infra | 03:06 | |
| *** bdodd has joined #openstack-infra | 03:31 | |
| *** vishalmanchanda has joined #openstack-infra | 04:38 | |
| *** ociuhandu has joined #openstack-infra | 04:40 | |
| *** ociuhandu has quit IRC | 04:45 | |
| *** soniya29 is now known as soniya29|sick | 04:56 | |
| *** ykarel has joined #openstack-infra | 05:23 | |
| *** evrardjp has quit IRC | 05:33 | |
| *** evrardjp has joined #openstack-infra | 05:33 | |
| *** rcernin has quit IRC | 05:43 | |
| *** rcernin has joined #openstack-infra | 05:55 | |
| *** matt_kosut has joined #openstack-infra | 06:10 | |
| *** zbr has quit IRC | 06:13 | |
| *** zbr has joined #openstack-infra | 06:14 | |
| *** hamalq has quit IRC | 06:17 | |
| *** hamalq has joined #openstack-infra | 06:18 | |
| *** ociuhandu has joined #openstack-infra | 06:28 | |
| *** gyee has quit IRC | 06:30 | |
| *** ociuhandu has quit IRC | 06:33 | |
| *** sboyron has joined #openstack-infra | 06:42 | |
| *** lmiccini has joined #openstack-infra | 06:49 | |
| *** sboyron has quit IRC | 06:55 | |
| *** ociuhandu has joined #openstack-infra | 07:02 | |
| *** ociuhandu has quit IRC | 07:06 | |
| *** zzzeek has quit IRC | 07:13 | |
| *** zzzeek has joined #openstack-infra | 07:14 | |
| *** eolivare has joined #openstack-infra | 07:27 | |
| *** lpetrut has joined #openstack-infra | 07:38 | |
| *** rcernin has quit IRC | 07:38 | |
| *** rcernin has joined #openstack-infra | 07:44 | |
| *** rcernin has quit IRC | 07:48 | |
| *** slaweq has joined #openstack-infra | 07:56 | |
| *** zxiiro has quit IRC | 07:57 | |
| *** hashar has joined #openstack-infra | 08:00 | |
| *** rcernin has joined #openstack-infra | 08:01 | |
| *** hashar has quit IRC | 08:04 | |
| *** hashar has joined #openstack-infra | 08:04 | |
| *** sboyron has joined #openstack-infra | 08:05 | |
| *** mmethot_ has joined #openstack-infra | 08:07 | |
| *** ociuhandu has joined #openstack-infra | 08:08 | |
| *** ociuhandu has quit IRC | 08:08 | |
| *** mmethot has quit IRC | 08:09 | |
| *** ociuhandu has joined #openstack-infra | 08:09 | |
| *** jcapitao has joined #openstack-infra | 08:12 | |
| *** ociuhandu has quit IRC | 08:15 | |
| *** rcernin has quit IRC | 08:16 | |
| *** ociuhandu has joined #openstack-infra | 08:22 | |
| *** ociuhandu has quit IRC | 08:27 | |
| *** rpittau|afk is now known as rpittau | 08:28 | |
| *** rcernin has joined #openstack-infra | 08:32 | |
| *** andrewbonney has joined #openstack-infra | 08:34 | |
| *** ociuhandu has joined #openstack-infra | 08:39 | |
| *** zbr has quit IRC | 08:44 | |
| *** ociuhandu has quit IRC | 08:45 | |
| *** zbr has joined #openstack-infra | 08:45 | |
| *** ociuhandu has joined #openstack-infra | 08:45 | |
| *** rcernin has quit IRC | 08:46 | |
| *** zbr has quit IRC | 08:47 | |
| *** zbr has joined #openstack-infra | 08:47 | |
| *** zbr has quit IRC | 08:49 | |
| *** tkajinam has quit IRC | 08:49 | |
| *** zbr has joined #openstack-infra | 08:49 | |
| *** zbr has quit IRC | 08:52 | |
| *** zbr has joined #openstack-infra | 08:52 | |
| *** jpena|off is now known as jpena | 08:57 | |
| *** lucasagomes has joined #openstack-infra | 09:02 | |
| *** tosky has joined #openstack-infra | 09:05 | |
| *** derekh has joined #openstack-infra | 09:07 | |
| *** sshnaidm is now known as sshnaidm|ruck | 09:34 | |
| *** xek has joined #openstack-infra | 09:36 | |
| *** rcernin has joined #openstack-infra | 09:42 | |
| *** rcernin has quit IRC | 09:42 | |
| *** ociuhandu has quit IRC | 09:49 | |
| *** carloss has joined #openstack-infra | 09:54 | |
| *** sshnaidm|ruck has quit IRC | 10:01 | |
| *** sshnaidm has joined #openstack-infra | 10:01 | |
| *** happyhemant has joined #openstack-infra | 10:05 | |
| *** ociuhandu has joined #openstack-infra | 10:16 | |
| *** ociuhandu has quit IRC | 10:16 | |
| *** gfidente has joined #openstack-infra | 10:21 | |
| openstackgerrit | Merged openstack/project-config master: Remove openstack-python and openstack-docs job for x/vmtp project. https://review.opendev.org/c/openstack/project-config/+/764054 | 10:22 |
|---|---|---|
| *** ociuhandu has joined #openstack-infra | 10:26 | |
| *** ykarel_ has joined #openstack-infra | 10:27 | |
| *** ykarel_ is now known as ykarel|lunch | 10:28 | |
| *** ykarel has quit IRC | 10:29 | |
| *** hashar is now known as hasharLunch | 10:32 | |
| *** icey_ has joined #openstack-infra | 10:43 | |
| *** icey has quit IRC | 10:44 | |
| *** ociuhandu has quit IRC | 10:46 | |
| *** ociuhandu has joined #openstack-infra | 10:46 | |
| *** ociuhandu has quit IRC | 10:48 | |
| *** ociuhandu has joined #openstack-infra | 10:48 | |
| *** ramishra_ has joined #openstack-infra | 10:54 | |
| *** sshnaidm is now known as sshnaidm|afk | 10:55 | |
| *** ramishra has quit IRC | 10:57 | |
| *** ysandeep is now known as ysandeep|afk | 11:04 | |
| *** dtantsur|afk is now known as dtantsur | 11:19 | |
| *** ykarel|lunch is now known as ykarel | 11:30 | |
| *** ykarel_ has joined #openstack-infra | 11:35 | |
| *** ykarel has quit IRC | 11:38 | |
| *** sshnaidm|afk is now known as sshnaidm|ruck | 11:39 | |
| *** icey_ has quit IRC | 11:44 | |
| *** icey has joined #openstack-infra | 11:58 | |
| *** ociuhandu has quit IRC | 12:02 | |
| *** iurygregory has joined #openstack-infra | 12:03 | |
| *** jcapitao is now known as jcapitao_lunch | 12:06 | |
| *** rfolcOUT has joined #openstack-infra | 12:06 | |
| *** rfolco has joined #openstack-infra | 12:07 | |
| *** ociuhandu has joined #openstack-infra | 12:23 | |
| *** ociuhandu has quit IRC | 12:27 | |
| *** ociuhandu has joined #openstack-infra | 12:29 | |
| zbr | clarkb: fungi: corvus what can we do to lower the amount of time reviews are stalling waiting for infra-cores? | 12:30 |
| *** dtantsur is now known as dtantsur|brb | 12:31 | |
| zbr | that is particularly an issue if you are on EU timezones as I doubt user can ping anyone on irc before ~2PM GMT. | 12:32 |
| *** ociuhandu has quit IRC | 12:33 | |
| *** jpena is now known as jpena|lunch | 12:36 | |
| *** vesper11 has quit IRC | 12:53 | |
| *** ociuhandu has joined #openstack-infra | 12:53 | |
| *** Goneri has joined #openstack-infra | 12:55 | |
| *** ociuhandu has quit IRC | 12:57 | |
| *** eharney has joined #openstack-infra | 13:01 | |
| *** hamalq has quit IRC | 13:01 | |
| *** jcapitao_lunch is now known as jcapitao | 13:03 | |
| *** hamalq has joined #openstack-infra | 13:05 | |
| *** ociuhandu has joined #openstack-infra | 13:10 | |
| *** rlandy has joined #openstack-infra | 13:18 | |
| *** ociuhandu has quit IRC | 13:22 | |
| *** ykarel_ is now known as ykarel | 13:30 | |
| *** ysandeep|afk is now known as ysandeep | 13:30 | |
| *** hasharLunch is now known as hashar | 13:43 | |
| *** dtantsur|brb is now known as dtantsur | 13:46 | |
| *** lbragstad has quit IRC | 13:50 | |
| *** jpena|lunch is now known as jpena | 13:51 | |
| *** ociuhandu has joined #openstack-infra | 13:53 | |
| fungi | zbr: ianw and frickler are usually around earlier than that | 13:55 |
| *** lbragstad has joined #openstack-infra | 13:55 | |
| zbr | fungi: i tried getting ianw on irc few times and was quite hard based on timestamps, I think he may in Asia (or working nights) | 13:57 |
| *** ociuhandu has quit IRC | 13:58 | |
| fungi | australia | 13:58 |
| fungi | and yeah we're a little thin in between with frickler on vacation at the moment | 13:59 |
| fungi | is there something urgent/broken? i'm just catching up on e-mail/scrollback now and about to go into a conference call meeting but can take a look shortly | 13:59 |
| *** ociuhandu has joined #openstack-infra | 14:02 | |
| zbr | sadly is an endemic issue, i do have examples but i doubt they count as urgent. | 14:05 |
| zbr | still, if various changes pile up, i endup loosing interest in doing anything on repos which have stalling PRs (not counting those with requested changes) | 14:05 |
| *** ociuhandu has quit IRC | 14:06 | |
| zbr | also the timing makes it harder because i mostly have 2-4h a day when i may be able to engage on irc with a core reviewer. | 14:06 |
| zbr | i have no idea what is the best way to addresses it, but i have the impression that it would not sort-itself out naturally | 14:07 |
| *** dchen has quit IRC | 14:09 | |
| zbr | maybe clarkb or corvus can help with https://review.opendev.org/c/opendev/gerritlib/+/729966 | 14:09 |
| *** auristor has quit IRC | 14:10 | |
| fungi | zbr: given you seem to have addressed clarkb's most recent question in that (as well as his and corvus's earlier questions) i'm going ahead and approving it | 14:16 |
| zbr | https://review.opendev.org/c/opendev/gerritlib/+/754784 should be no brainer, removes platforms which we didn't even test. | 14:19 |
| *** ociuhandu has joined #openstack-infra | 14:49 | |
| *** ociuhandu has quit IRC | 14:51 | |
| *** ociuhandu has joined #openstack-infra | 14:52 | |
| *** zxiiro has joined #openstack-infra | 14:53 | |
| *** xek has quit IRC | 14:57 | |
| *** xek has joined #openstack-infra | 14:58 | |
| *** ociuhandu has quit IRC | 14:58 | |
| *** hashar is now known as hasharAway | 15:05 | |
| zbr | basepython = python3 is quite dangerous entry to have in tox.ini, it confuses tox into user other interpreters even when you do "tox -e py3X" | 15:05 |
| zbr | the tools warns about it, but continues to run with mismatching interpreter. | 15:06 |
| zbr | fungi: thanks for gerritlib merge, now i can run the tests locally on any version of python. | 15:09 |
| *** auristor has joined #openstack-infra | 15:09 | |
| fungi | zbr: yes, "basepython = python3" does unexpected things for pyXY testenvs, unless you also add "ignore_basepython_conflict = True" in the global [tox] section (supposedly that will eventually become the default behavior after a transitional period) | 15:12 |
| *** ociuhandu has joined #openstack-infra | 15:12 | |
| fungi | https://tox.readthedocs.io/en/latest/config.html#conf-ignore_basepython_conflict | 15:13 |
| zbr | yeah, i am reading it again, but i am not sure if wisdom is what i get by reading it again, these two paragraphs..... | 15:15 |
| zbr | i personally do not see the use of that option, as someone that very often specifies explicit python versions when I run tests. | 15:16 |
| fungi | i agree the explanation for that option in their documentation is lacking clarity. in short it's the fix they implemented for the bug we raised about that confusing behavior | 15:16 |
| zbr | I do find tox behaving as expected only when both entries are missing from tox.ini files. | 15:17 |
| fungi | to rephrase the explanation, ignore_basepython_conflict allows you to continue to get python 3.8 when you tox -e py38 even if you have basepython for your testenv set to something else like python3.7 | 15:17 |
| fungi | https://codesearch.opendev.org/?q=ignore_basepython_conflict | 15:18 |
| fungi | we've relied on it heavily in the python 2->3 transition | 15:18 |
| *** ociuhandu has quit IRC | 15:20 | |
| zbr | and the cost of breaking when someone added basepython=python3.6 on a specific environment. | 15:20 |
| zbr | there are cases when someone wants to run an environment with only a particular version of python, linting could be such case. | 15:20 |
| zbr | as it is know that some linters give different results based on what version of python you use, even if your project may support a wide range of pythons. | 15:21 |
| zbr | i know, about =python3 as a hack during transition, added it myself to many repos. | 15:21 |
| zbr | but imho, python3 was an unversioned/ubdetermined python3. | 15:22 |
| *** ysandeep is now known as ysandeep|away | 15:23 | |
| zbr | i guess that a better approach is to use "py3" for those that do not care about version, and avoid using ignore_basepython_conflict and base-python. | 15:24 |
| *** ociuhandu has joined #openstack-infra | 15:26 | |
| fungi | i'm not sure what you're asserting breaks | 15:26 |
| fungi | if you create an explicit [testenv:py38] and then set basepython = python3.7 inside that, yes i'm not sure exactly what it will do, that might still use python3.8 instead | 15:27 |
| fungi | when ignore_basepython_conflict is set true | 15:27 |
| *** kmarc has joined #openstack-infra | 15:27 | |
| fungi | it seems like an unusual corner case though, rare enough that you can probably afford other workarounds for a project which needs to do something that strange | 15:28 |
| zbr | yeah, just dicovered this 5s ago. a lot of magic behind tox. | 15:28 |
| zbr | in old versions of tox using ignore_basepython_conflict was recipe for disaster (getting CI to report passed for unexpected version of python) | 15:29 |
| *** dklyle has joined #openstack-infra | 15:29 | |
| zbr | fungi: do you recommend including only "py3" in envlist or full list of supported versions? | 15:30 |
| fungi | i thought it was widely known in openstack circles for the past couple of years. it came up when we first started trying to switch the default python away from 2.7 and we raised it as a bug at the time. they gave us ignore_basepython_conflict in tox 3.1.0 so we could have a cleaner solution than making explicit [testenv:pyXY] entries just so we could override basepython differently in each one | 15:30 |
| *** ociuhandu has quit IRC | 15:31 | |
| fungi | zbr: i'm an advocate of keeping envlist limited to py3 rather than sticking py36,py37,py38,py39 in there for a few reasons: 1. most devs are unlikely to have every one of those interpreter versions installed, 2. local testing with one python version is generally good enough and having the ci system catch your very version-specific bug for you instead isn't the end of the world, 3. it avoids needing | 15:32 |
| fungi | to remember to constantly edit the envlist line every time your supported interpreter versions change | 15:32 |
| fungi | also 4. it's not hard to explicitly tox -e py38 if your default python3 is 3.7 but you know you have 3.8 installed and you want to specifically test with it | 15:33 |
| zbr | i have to agree the keeping the list maintained is a pita and because i can still mention explicit version, is ok. if i see ci to fail with specific version, I can run with that locally. | 15:33 |
| zbr | i do happen to use pyenv and have an insane versions of pythons, but i am sure that is not the average dev experience | 15:34 |
| fungi | my opinion is that tox is also not the right place to declare what python versions your project supports, that's why we have package metadata | 15:34 |
| zbr | yep, but tox declares the test matrix, and including all versions in the test matrix is usually a good idea. | 15:35 |
| fungi | and yeah, that's more or less what i do too (i don't use pyenv, i just make each python from the git source tree and make altinstall to a subpath of my homedir, then deep-link from ~/bin into those) | 15:35 |
| zbr | maybe gabor will add a feature to get the list from setupcfg | 15:35 |
| fungi | would probably want to parse both setup.cfg (that's not just a pbr thing, setuptools supports it too) and pyproject.toml (presumptive successor) | 15:36 |
| zbr | apparently on gerritbot stestr brings a perverse amount of deprecation warnings from subunit, which seams unmaintained. | 15:41 |
| *** lpetrut has quit IRC | 15:41 | |
| zbr | in fact i see unittest2 being installed which should never happen on py3 projects. | 15:41 |
| fungi | might be worth checking with mtreinish on those | 15:49 |
| fungi | but fair warning, you're not going to be able to get rid of every last deprecation warning on newer pythons until https://github.com/pypa/setuptools/issues/2086 is fixed in a new release of setuptools and the vendored copy in Python 3.8 and later updated with it | 15:50 |
| fungi | also probably not until python 3.6 is dropped entirely since it's unlikely the setuptools in it will ever be updated to a version which calls importlib instead of imp | 15:51 |
| fungi | or, alternatively, getting rid of setuptools entirely i guess (that's been on the pbr to do list for a while) | 15:51 |
| fungi | but yeah, at the moment even the latest versions of setuptools still call lib2to3 so that's going to raise deprecation warnings until it's addressed | 15:53 |
| *** slaweq has quit IRC | 15:54 | |
| clarkb | zbr: I'm not fully caught up on scrollback but helping us enables us to help you. I'm not sure if you've noticed but the last couple of months have been incredibly busy for infra root for a number of reasons | 15:54 |
| clarkb | I don't have time to do reviews on most things right now | 15:55 |
| zbr | sadly, the user-lookup in gerrit seams unusable, like 10s+ to return values. | 15:55 |
| clarkb | also the constant complaining is really making it hard for me to want to bother | 15:55 |
| clarkb | because it feels like the end result of all this work is that everyone is just unhappy and unwilling to help | 15:56 |
| fungi | zbr: what "user lookup"? the rest api? it's returning quickly for me | 15:56 |
| *** slaweq has joined #openstack-infra | 15:56 | |
| fungi | also system load on the server doesn't seem to be as bad today: http://cacti.openstack.org/cacti/graph.php?action=view&local_graph_id=26&rra_id=all | 15:56 |
| zbr | lookup is usually the term used for UI elements that perform server side searches, when you try to add someone to a review. | 15:56 |
| fungi | still not great, but maybe the fix to where we disabled autogc in the correct file had some positive impact | 15:57 |
| zbr | there performance seams to vary a lot from moment to moment, sometimes is <1s, sometimes nothing. | 15:57 |
| fungi | yeah, the server is still struggling under heavy load at peak utilization since the upgrade. we're trying lots of options to get it under control | 15:57 |
| zbr | i think that we need to assure that user lists are always in cache as this is very common activity: to add reviewers. | 15:58 |
| fungi | zbr: i appreciate your desire to state the obvious, we've already tuned the accounts cache to be as large as gerrit seems to want to use | 15:58 |
| zbr | on the other hand, I have the impression that the auto-completer is smarter | 15:58 |
| *** ociuhandu has joined #openstack-infra | 15:59 | |
| fungi | currently, the accounts cache has a 99% hit ratio in memory and 100% hit ratio on disk (it's a disk-backed cache for that one) | 16:00 |
| zbr | maybe is something else, like reaching http server connection limits? | 16:01 |
| fungi | well, we've been keeping tabs on the apache mod_status scoreboard too | 16:04 |
| zbr | i wonder if there is a browser plugin which can be used to report average response time, having few users reporting it could make use identify bottlenecks much easier than on irc. | 16:05 |
| fungi | it's fairly idle at the moment, most workers are either in waiting for connection or open slot with no current process | 16:05 |
| zbr | long time ago i used new relic for something like that. | 16:05 |
| clarkb | I don't think we need more people reporting issues at thsi point, we have the logs and metrics we've been looking at and have been picking off one thing at a time | 16:05 |
| clarkb | if we exhaust our abilities doing ^ then we can start looking at other tools | 16:06 |
| clarkb | and so far it seems the work we have done has improved things | 16:06 |
| fungi | and yeah, next on the list is switching from java 8 to 11, which is supposed to help performance as well | 16:09 |
| fungi | then it's probably time to start looking at updating to gerrit 3.3 which also supposedly has some performance improvements | 16:10 |
| *** redrobot has quit IRC | 16:11 | |
| clarkb | ya, though at least one person had trouble upgraded on the mailing list yesterday (I helped them by pointing out the manual downgrade steps). I expect we'll want to do a fair bit of testing for that one (not 3.2 levels but enough to be happy with it first) | 16:11 |
| clarkb | in theory it is a very easy upgrade though, only one schema update and no index updates | 16:12 |
| zbr | while the attention-set seams like a good replacement for assignee, i doubt we want to jump so soon on 3.3, lets other pay the cost of being an early adopter. | 16:12 |
| zbr | but java 11, seams like best next approach, we cannot escape it anyway. | 16:13 |
| fungi | well, the difference between 3.2 and 3.3 isn't substantial, and if it fixes some of our performance issues it might be worth doing sooner so long as some cursory testing doesn't identify problems | 16:13 |
| fungi | they've supposedly also started supporting java 15 with gerrit 3.3, but that does seem like a risky jump given the bugs they've found | 16:14 |
| clarkb | fungi: jgit had problems i don't know if they fixed them yet on 15 | 16:15 |
| fungi | yeah, it's stuff like that which makes me squeamish about moving to openjdk 15 just yet | 16:16 |
| fungi | they apparently realized they were missing testing for it until very recently too | 16:16 |
| clarkb | luca reports running 3.2 on java 11 since day one at gerrithub so fairly confident in that update | 16:16 |
| mtreinish | fungi: unittest2 is still a dep for released testtools. It was a rolling backport from stdlib unittest but stopped backporting things around. It's still needed for python 3.4 testtools support, which the released testtools still supports | 16:20 |
| fungi | zbr: ^ maybe that helps you | 16:20 |
| fungi | thanks mtreinish! | 16:21 |
| mtreinish | the next release of testtools should drop that requirement, I just need another maintainer to review: https://github.com/testing-cabal/testtools/pull/277 | 16:21 |
| clarkb | mtreinish: o/ ohai its been a long time | 16:22 |
| clarkb | mtreinish: still making computers do crazy things at really cold tempuratures? | 16:22 |
| zbr | mtreinish: nice to see you here! any chance for a refresh on stestr deps to bring them more towards py36-py39 range? | 16:22 |
| zbr | with the risk of upsetting some around here, i am more of pytest user. | 16:23 |
| fungi | sounds like the testtools update needs to happen first | 16:23 |
| fungi | since stestr depends on it | 16:23 |
| clarkb | zbr: and as we're happy to point out the use of stestr which sticks to standards means you can use pytest if you choose to | 16:23 |
| mtreinish | clarkb: hey, yeah still doing the quantum computing stuff. It's been fun so far | 16:23 |
| clarkb | zbr: but if we use pytest then alternatives are unlikely to work due to its lack of adhering to standards | 16:24 |
| mtreinish | zbr: yeah what fungi said stestr depends on testtools for it's realtime results streaming. I've removed most of the dependency on testtools from stestr at this point, but that piece is still critical | 16:24 |
| zbr | i am aware of risks, while is possible to use pytest with pure unittest tests, its real power comes when you diverge from the standards. | 16:25 |
| fungi | sure, but diverging from standards also carries plenty of cost | 16:26 |
| zbr | what can we do about testools? it does not seam to actively maintained. | 16:27 |
| zbr | my impression is that the CI is broken and I open PRs to drop unittest2 1.5 years old. | 16:28 |
| mtreinish | it's not super active, I try my best to keep on top of it. But I have been busier with work projects than I was expecting when I agreed to be added as a maintainer | 16:28 |
| zbr | should we just fork it, fix what we care about and release that fork? | 16:29 |
| mtreinish | if I get sufficient reviews on #277 and once that merges I can prepare a new testtools release | 16:29 |
| mtreinish | but also instead of forking it, why not get some people to volunteer to work on it. I can discuss adding maintainers in the core group. I don't think there would be much opposition | 16:30 |
| fungi | yeah, if there are enough people to maintain a fork, then they could potentially be volunteers to just help maintain the original instead? | 16:31 |
| zbr | mtreinish: i already have a very long list of projects i nurture, so I can only help with packaging and fixes like these, not real development. | 16:31 |
| zbr | like fixing broken CI, fixing dependency conflict | 16:32 |
| * zbr hates forking, is last resort action. | 16:33 | |
| *** lmiccini has quit IRC | 16:35 | |
| zbr | mtreinish: my github id is ssbarnea, happy to help addressing these deprecations (hopefully we do not endup having to fix 50 other dependencies), | 16:35 |
| mtreinish | well is there anything big besides the unittest2 thing. I think that was the only big deprecation/compat issue with modern python versions | 16:57 |
| *** Guest18921 has joined #openstack-infra | 16:58 | |
| *** gyee has joined #openstack-infra | 16:58 | |
| *** Guest18921 is now known as redrobot | 17:01 | |
| *** ykarel is now known as ykarel|away | 17:04 | |
| *** rpittau is now known as rpittau|afk | 17:04 | |
| *** ykarel|away has quit IRC | 17:09 | |
| *** ykarel|away has joined #openstack-infra | 17:17 | |
| *** gbutnaru has joined #openstack-infra | 17:19 | |
| *** ykarel|away has quit IRC | 17:22 | |
| *** ociuhandu has quit IRC | 17:23 | |
| *** lucasagomes has quit IRC | 17:25 | |
| *** ociuhandu has joined #openstack-infra | 17:28 | |
| *** ociuhandu has quit IRC | 17:28 | |
| mtreinish | fungi, clarkb: this looks like a pbr issue: https://github.com/testing-cabal/testtools/pull/281/checks?check_run_id=1487744938#step:6:111 | 17:29 |
| mtreinish | it's running 'pip install .[test]' in the ci job and it looks like pip can't find the version correctly with the setuptools extra | 17:29 |
| mtreinish | and the new pip depsolver isn't happy with that | 17:30 |
| fungi | oh, right, i'm always confused by github's requirements to sign into my account there before i can look at public ci logs | 17:30 |
| fungi | ERROR: Cannot install testtools and testtools[test]==0.0.0 because these package versions have conflicting dependencies. | 17:31 |
| fungi | that? | 17:31 |
| fungi | looks almost certainly related to this week's pip release which enables the new dep solver | 17:32 |
| fungi | i'm not sure it's a pbr-specific problem, looks more like whatever test harness is trying to install the project into the venv | 17:33 |
| fungi | mtreinish: i'm not super familiar with github workflows, but i think this is the problem: https://github.com/testing-cabal/testtools/blob/master/.github/workflows/test.yml#L45 | 17:36 |
| fungi | i think new pip won't allow you to install both testtools and testtools[extras] | 17:37 |
| fungi | though i can't quite tell where testtools by itself is getting installed | 17:38 |
| mtreinish | I don't think it installs testtools by itself | 17:38 |
| *** artom has quit IRC | 17:38 | |
| mtreinish | the only line that is installing testtools is l45 there | 17:38 |
| mtreinish | the weirdness comes in with the extra, because pretty much everything in there depends on testtools: https://github.com/testing-cabal/testtools/blob/master/setup.cfg#L27 | 17:39 |
| fungi | oh, that's what it is, i bet | 17:39 |
| fungi | so yeah it wants to install testtools as a dependency of the things in testtools[extra] | 17:40 |
| fungi | resulting in a conflicting circular dependency | 17:40 |
| fungi | i guess the ==0.0.0 there is the real problem then? the version metadata from the git repo isn't being applied? | 17:41 |
| mtreinish | yeah, that's what I was thinking | 17:42 |
| fungi | i'll experiment and see if i can reproduce in a straightforward way | 17:42 |
| *** eolivare has quit IRC | 17:44 | |
| *** sshnaidm|ruck is now known as sshnaidm|afk | 17:47 | |
| fungi | it's not as simple as just installing the master branch tip's [test] extra into a venv | 17:49 |
| fungi | Successfully installed argparse-1.4.0 extras-1.0.0 fixtures-3.0.0 linecache2-1.0.0 pbr-5.5.1 python-mimeparse-1.6.0 testresources-2.0.1 testscenarios-0.5.0 testtools-2.4.1.dev24 traceback2-1.4.0 unittest2-1.1.0 | 17:54 |
| *** jcapitao has quit IRC | 17:54 | |
| fungi | so at least in my local testing pbr seems to be supplying a version number based on the state of the git repo | 17:54 |
| mtreinish | oh, I wonder if it's a shallow clone issue. I think by default gha does a shallow clone of the checkout | 17:55 |
| mtreinish | so I bet the tags aren't there | 17:55 |
| fungi | that's what i'm starting to suspect | 17:55 |
| fungi | it's acting like there's no git for pbr to reference | 17:55 |
| fungi | i'm seeing if blowing away .git gives me similar behavior | 17:56 |
| fungi | okay, not quite... the error condition is definitely different when pbr has no .git | 17:58 |
| fungi | i'll test the limited clone theory | 17:58 |
| fungi | so if i clone with --no-tags i get the same error | 18:00 |
| fungi | mtreinish: and if i git pull --tags then it works | 18:01 |
| fungi | so my guess would be github actions is setting up the clone with --no-tags or equivalent | 18:01 |
| *** jpena is now known as jpena|off | 18:02 | |
| fungi | or git fetch --tags is probably safer for this case | 18:02 |
| fungi | definitely lmk if that doesn't make it go and i can keep digging? | 18:03 |
| mtreinish | thanks | 18:06 |
| mtreinish | yeah I'll push a PR to update the job definition, there is an option to just do a full clone (I've had to do that in other jobs before that need the full history, like with reno) | 18:06 |
| *** lpetrut has joined #openstack-infra | 18:09 | |
| fungi | oh, good point, a shallow clone won't have the objects for the tags to correspond to, so yeah you might need at least the full history of that branch | 18:11 |
| *** derekh has quit IRC | 18:14 | |
| *** hasharAway is now known as hashar | 18:15 | |
| zbr | the issue is not pbr specific, i got the same issue with setuptools-scm when lacking full clone. | 18:16 |
| zbr | but it could prove as a good idea to attempt to go a git pull tags from inside pbr before failing. | 18:16 |
| *** andrewbonney has quit IRC | 18:19 | |
| fungi | well, that may not help if the clone is shallow as mtreinish suggests, in which case you really do need to clone differently | 18:20 |
| mtreinish | well it was an issue with reno, but the full history matters more there. Just having the tag metadata might be enough I haven't tried it | 18:24 |
| zbr | if so is right, it may be possible https://stackoverflow.com/questions/6802145/how-to-convert-a-git-shallow-clone-to-a-full-clone | 18:24 |
| zbr | imho only tags matter for, but it may be interesting to make some tests, maybe we can avoid the 0.0.0 fallback for some users. | 18:25 |
| fungi | oh, neat, i didn't realize you could adjust the depth when fetching | 18:25 |
| fungi | and missed that --unshallow became a fetch option too | 18:26 |
| zbr | sadly I did not see get me last 2 tags. | 18:26 |
| *** gfidente is now known as gfidente|afk | 18:40 | |
| *** dtantsur is now known as dtantsur|afk | 18:47 | |
| *** vishalmanchanda has quit IRC | 18:58 | |
| *** ramishra_ has quit IRC | 19:00 | |
| *** artom has joined #openstack-infra | 19:15 | |
| *** iurygregory has quit IRC | 19:29 | |
| *** dhill has quit IRC | 20:02 | |
| *** dhill has joined #openstack-infra | 20:07 | |
| *** vesper11 has joined #openstack-infra | 20:15 | |
| *** vesper has joined #openstack-infra | 20:19 | |
| *** vesper11 has quit IRC | 20:19 | |
| *** hashar has quit IRC | 20:23 | |
| *** lpetrut has quit IRC | 20:29 | |
| *** dhill has quit IRC | 20:53 | |
| *** vesper has quit IRC | 21:06 | |
| *** raissa has joined #openstack-infra | 21:15 | |
| *** dhill has joined #openstack-infra | 21:18 | |
| *** rfolco has quit IRC | 21:20 | |
| *** iurygregory has joined #openstack-infra | 21:28 | |
| *** vesper11 has joined #openstack-infra | 21:30 | |
| *** sboyron has quit IRC | 21:43 | |
| *** raissa has quit IRC | 21:45 | |
| *** gbutnaru has quit IRC | 22:06 | |
| *** matt_kosut has quit IRC | 22:27 | |
| *** matt_kosut has joined #openstack-infra | 22:28 | |
| *** matt_kosut has quit IRC | 22:33 | |
| *** rcernin has joined #openstack-infra | 22:34 | |
| *** ajitha has quit IRC | 22:35 | |
| *** tkajinam has joined #openstack-infra | 23:00 | |
| *** slaweq has quit IRC | 23:02 | |
| *** rlandy has quit IRC | 23:23 | |
| *** dchen has joined #openstack-infra | 23:37 | |
| *** Goneri has quit IRC | 23:50 | |
| *** happyhemant has quit IRC | 23:54 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!