| *** dangtrinhnt has joined #openstack-requirements | 00:03 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/requirements master: update constraint for ldappool to new release 2.3.1 https://review.openstack.org/614052 | 00:40 |
|---|---|---|
| tonyb | prometheanfire: Yes they are pipenv esp. They'll help us some | 00:59 |
| prometheanfire | tonyb: I haven't seen a way that pipfiles are better than requirements/constraints files quite yet | 01:00 |
| tonyb | prometheanfire: They're not a whole lot better but they do allow for a little more mixing than we have today | 01:02 |
| prometheanfire | mixing? | 01:05 |
| prometheanfire | pipfiles don't allow python mixing, but keeping separate files would 'fix' that | 01:07 |
| tonyb | Sorry I was speaking of pipenv not pipfiles | 01:11 |
| prometheanfire | ah | 01:12 |
| prometheanfire | pipenv install -r global-requirements.txt | 01:12 |
| prometheanfire | should be fun | 01:13 |
| prometheanfire | I heard it does depsolving | 01:13 |
| prometheanfire | I kinda want to see if I can get it to print a graph | 01:14 |
| prometheanfire | just for the presentation | 01:14 |
| prometheanfire | I should have time'd it | 01:15 |
| *** hongbin has joined #openstack-requirements | 01:23 | |
| tonyb | hehe | 01:25 |
| prometheanfire | lolnope | 01:26 |
| prometheanfire | Warning: Your dependencies could not be resolved. You likely have a mismatch in your sub-dependencies. | 01:26 |
| prometheanfire | First try clearing your dependency cache with $ pipenv lock --clear, then try the original command again. | 01:27 |
| prometheanfire | Alternatively, you can use $ pipenv install --skip-lock to bypass this mechanism, then run $ pipenv graph to inspect the situation. | 01:27 |
| prometheanfire | Hint: try $ pipenv lock --pre if it is a pre-release dependency. | 01:27 |
| prometheanfire | Could not find a version that matches dogpile-cache | 01:27 |
| prometheanfire | No versions found | 01:27 |
| prometheanfire | hmm, rate limited? | 01:27 |
| tonyb | No idea, perhaps | 01:28 |
| * prometheanfire shrugs | 01:31 | |
| prometheanfire | retry, it needs more sucessorator | 01:31 |
| prometheanfire | well, was worth a trypipenv.patched.notpip._internal.exceptions.InstallationError: Command "python setup.py egg_info" failed with error code 1 in /tmp/tmp0z_99leabuild/qpid-python/ | 01:56 |
| prometheanfire | pipenv.patched.notpip._internal.exceptions.InstallationError: Command "python setup.py egg_info" failed with error code 1 in /tmp/tmp0z_99leabuild/qpid-python/ | 01:56 |
| *** udesale has joined #openstack-requirements | 03:49 | |
| *** hongbin has quit IRC | 04:23 | |
| *** ccamacho has quit IRC | 04:26 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/requirements master: Updated from generate-constraints https://review.openstack.org/614078 | 06:18 |
| openstackgerrit | Merged openstack/requirements master: update constraint for os-win to new release 4.1.1 https://review.openstack.org/613959 | 06:37 |
| *** dangtrinhnt has quit IRC | 06:43 | |
| *** dangtrinhnt has joined #openstack-requirements | 06:48 | |
| *** tonyb has quit IRC | 06:50 | |
| openstackgerrit | Merged openstack/requirements master: Updated from generate-constraints https://review.openstack.org/613845 | 06:55 |
| *** ccamacho has joined #openstack-requirements | 08:04 | |
| *** jpich has joined #openstack-requirements | 08:26 | |
| *** e0ne has joined #openstack-requirements | 08:41 | |
| *** ccamacho has quit IRC | 09:01 | |
| *** ccamacho has joined #openstack-requirements | 10:11 | |
| *** e0ne has quit IRC | 10:32 | |
| *** e0ne has joined #openstack-requirements | 10:56 | |
| *** udesale has quit IRC | 11:24 | |
| *** vpickard_ is now known as vpickard | 11:39 | |
| *** vpickard is now known as vpickard_ | 11:39 | |
| *** dtantsur|afk is now known as dtantsur | 11:50 | |
| *** e0ne has quit IRC | 12:21 | |
| *** zul has quit IRC | 12:27 | |
| *** vpickard_ is now known as vpickard | 12:29 | |
| *** e0ne has joined #openstack-requirements | 13:08 | |
| *** udesale has joined #openstack-requirements | 13:08 | |
| openstackgerrit | Matthew Thode proposed openstack/requirements master: Updated from generate-constraints https://review.openstack.org/614078 | 13:32 |
| prometheanfire | dhellmann: isn't that it | 13:53 |
| dhellmann | prometheanfire : ? | 13:54 |
| prometheanfire | dhellmann: doesn't that it's installing and using twine in an openstack project mean that it should be in reqs? | 13:54 |
| dhellmann | oh, no. the job builds its own virtualenv | 13:54 |
| dhellmann | twine does not need to be co-installed with openstack | 13:55 |
| dhellmann | nothing else needs to be installed to build a package | 13:55 |
| prometheanfire | right, but it is needed to upload, I thought that was why we used twine | 13:55 |
| dhellmann | that job also does its own installation | 13:56 |
| prometheanfire | should that install be using reqs/constraints then? | 13:56 |
| dhellmann | why? | 13:56 |
| prometheanfire | it was a question :P | 13:56 |
| prometheanfire | smcginnis: ^ | 13:56 |
| dhellmann | why would it? | 13:56 |
| dhellmann | if we need to constrain it, we can do that in the step that installs the package in the job | 13:56 |
| prometheanfire | I think all things should be in reqs/constraints more or less | 13:57 |
| dhellmann | I think that list should be limited to the set of things we need to be co-installable for production-like installations and test jobs | 13:58 |
| dhellmann | that's why we blacklist things like linters | 13:58 |
| openstackgerrit | Monty Taylor proposed openstack/requirements master: update constraint for openstacksdk to new release 0.18.1 https://review.openstack.org/612474 | 13:59 |
| prometheanfire | true, though the reason I saw for blacklisting linters was because diferent projects had VASTLY diferent versions of the linters | 13:59 |
| dhellmann | mordred : we're about to release 0.19.0, right? don't you want that? | 13:59 |
| prometheanfire | gentoo co-installs and runs linters with the other tests at times | 14:00 |
| dhellmann | prometheanfire : right, because they don't need to be the same. Projects don't need to control which version of twine they use. They *must* work with the latest version. | 14:00 |
| prometheanfire | dhellmann: if projects know this, then ok | 14:00 |
| dhellmann | prometheanfire : it doesn't make any sense to me for a distro to run a linter check | 14:00 |
| prometheanfire | same class as setuptools/pip | 14:00 |
| dhellmann | apparently projects didn't understand that, so we need to communicate it better | 14:00 |
| mordred | dhellmann: I do ... the constraints patch above is actually to test to see if a cliff patch fixes a failure that was being seen | 14:00 |
| dhellmann | prometheanfire : right | 14:00 |
| prometheanfire | dhellmann: oh, I know, didn't say I agreed with that | 14:00 |
| dhellmann | mordred : ah, ok | 14:00 |
| mordred | dhellmann: https://review.openstack.org/#/c/614218/ fwiw | 14:01 |
| mordred | dhellmann: http://logs.openstack.org/74/612474/2/check/cross-osc-build-sphinx-docs/42250a7/job-output.txt.gz#_2018-10-26_08_02_08_597700 is the log failure | 14:01 |
| dhellmann | mordred : lgtm | 14:01 |
| mordred | cool. it feels dirty - but I guess that's the best way to test if something is a class | 14:02 |
| dhellmann | adding some info about what caused the failure to the commit message would be good | 14:02 |
| mordred | would be nice if there was an 'isclass' | 14:02 |
| mordred | kk | 14:02 |
| dhellmann | you could do issubclass of meta | 14:02 |
| dhellmann | or of type | 14:02 |
| mordred | but you can't - because you can't pass a non-class to issubclass | 14:02 |
| smcginnis | prometheanfire: I didn't realize dhellmann's job covered things, so I can abandon the twine patch. | 14:02 |
| dhellmann | mordred : https://docs.python.org/3/library/inspect.html#inspect.isclass | 14:02 |
| mordred | dhellmann: I mean, I guess you could do that and trap for an exception | 14:02 |
| dhellmann | I'm not sure what version added that | 14:02 |
| smcginnis | That was only needed to be able to add to project's test-requirements, but those can just be removed now. | 14:03 |
| prometheanfire | smcginnis: k, thanks | 14:03 |
| mordred | dhellmann: it's in 2.7 - I'll use that | 14:03 |
| dhellmann | ++ | 14:03 |
| prometheanfire | smcginnis: dhellmann it may be needed in blacklist then, like the linters (with a comment as to why (same class as setuptools/pip)) | 14:03 |
| dhellmann | prometheanfire : we do not want projects adding twine to their requirements or test requirements lists, so keeping it out of the blacklist will cause an error if they try to add it | 14:04 |
| dhellmann | when this all-hands about the ibm acq is done I will post to the mailing list if smcginnis hasn't already done so | 14:04 |
| mordred | dhellmann: updated that patch | 14:05 |
| smcginnis | dhellmann: I'll let you post to the ML if you want. Otherwise I was just going to put up patches to remove it from projects jobs with an explanation in the commit message. | 14:05 |
| prometheanfire | dhellmann: ok, smcginnis mentioned adding it to test-reqs, which is why I ased | 14:05 |
| smcginnis | I would think if a project had not added it by now, they aren't planning on it. | 14:05 |
| mordred | dhellmann: so .... re: twine | 14:06 |
| dhellmann | smcginnis : when you have one or two of those done I can link to them, but I think it's useful to explain why so that no one comes along and tries to do this again later | 14:06 |
| mordred | python setup.py check -r -s is now throwing a warning telling people to use 'twine check' instaed | 14:07 |
| smcginnis | mordred: Yeah, that's what led to this. | 14:07 |
| dhellmann | yeah | 14:07 |
| mordred | cool gotcha. it's not the same, btw - I tried using it yesterday and it wants a dist to be made | 14:07 |
| smcginnis | mordred: Current recommendation is to use twine, but projects don't need it in their own tox.ini now because dhellmann has it coverd in a zuul job. | 14:07 |
| mordred | ah - awesome! | 14:07 |
| mordred | that's better - because I've got tox set up to skip sdist, which would make twine check annoying in tox :) | 14:08 |
| * mordred is late to the party, but happy to be here | 14:08 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/requirements master: update constraint for python-searchlightclient to new release 1.4.0 https://review.openstack.org/614222 | 14:11 |
| dhellmann | mordred : you're fashionably late. we've only been talking about it for < 1 hr | 14:12 |
| *** kiennt26 has joined #openstack-requirements | 14:21 | |
| dhellmann | smcginnis : I have an email ready to send, do you have a patch or gerrit topic for me to link to or do you want to link to my email in your commit messages? | 14:31 |
| *** e0ne has quit IRC | 14:33 | |
| *** e0ne has joined #openstack-requirements | 14:37 | |
| smcginnis | dhellmann: I haven't been able to get to it yet. I can link to the email in the commit messages. | 14:38 |
| dhellmann | smcginnis : wfm | 14:38 |
| smcginnis | dhellmann: I did see a few that have that in their linter jobs but do not have the release job in project-config. | 14:39 |
| smcginnis | So probably good to mention that in the ML thread. | 14:39 |
| dhellmann | smcginnis : http://lists.openstack.org/pipermail/openstack-dev/2018-October/136136.html | 14:40 |
| dhellmann | I probably should have tagged that release, too | 14:40 |
| dhellmann | oh, that's an interesting case | 14:41 |
| dhellmann | which projects don't have the release job? | 14:41 |
| smcginnis | dhellmann: I only took at the first two that came up in codesearch so far - monasca-notification and ansible-role-chrony. | 14:42 |
| dhellmann | ok | 14:42 |
| dhellmann | ansible-role-chrony seems to be using the template | 14:44 |
| dhellmann | as does monasca-notification | 14:44 |
| smcginnis | Yeah, just realized I missed that template. | 14:47 |
| *** dtantsur is now known as dtantsur|brb | 14:50 | |
| *** kiennt26 has quit IRC | 14:51 | |
| *** e0ne has quit IRC | 15:25 | |
| *** kiennt26 has joined #openstack-requirements | 15:31 | |
| *** e0ne has joined #openstack-requirements | 15:35 | |
| *** ccamacho has quit IRC | 15:48 | |
| *** kiennt26 has quit IRC | 15:49 | |
| *** dtantsur|brb is now known as dtantsur | 16:26 | |
| *** e0ne has quit IRC | 16:30 | |
| *** udesale has quit IRC | 16:43 | |
| *** jrist has quit IRC | 16:51 | |
| *** jrist has joined #openstack-requirements | 16:54 | |
| *** jpich has quit IRC | 17:08 | |
| *** ccamacho has joined #openstack-requirements | 17:23 | |
| openstackgerrit | Matthew Thode proposed openstack/requirements master: update future https://review.openstack.org/614314 | 17:40 |
| *** hongbin has joined #openstack-requirements | 17:42 | |
| hongbin | prometheanfire: hi, i replied your comment in https://review.openstack.org/#/c/609018/ , could you take a look once you have a chance? | 17:43 |
| *** dtantsur is now known as dtantsur|afk | 17:57 | |
| *** e0ne has joined #openstack-requirements | 18:10 | |
| prometheanfire | hongbin: was waiting for the reviews mentioned to merge, looks good now | 18:20 |
| hongbin | prometheanfire: ack | 18:20 |
| openstackgerrit | Merged openstack/requirements master: update constraint for ldappool to new release 2.3.1 https://review.openstack.org/614052 | 18:22 |
| *** e0ne has quit IRC | 18:25 | |
| *** e0ne has joined #openstack-requirements | 19:13 | |
| *** e0ne has quit IRC | 21:18 | |
| *** e0ne has joined #openstack-requirements | 22:49 | |
| *** tonyb has joined #openstack-requirements | 23:03 | |
| *** e0ne has quit IRC | 23:13 | |
| *** hongbin has quit IRC | 23:57 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!