openstackgerrit | Tony Breeds proposed openstack/requirements master: Add sphinx_feature_classification as a requirement https://review.openstack.org/481294 | 00:07 |
---|---|---|
*** odyssey4me has quit IRC | 00:53 | |
*** odyssey4me has joined #openstack-requirements | 00:53 | |
*** hongbin has joined #openstack-requirements | 00:57 | |
openstackgerrit | Merged openstack/requirements master: Updated from generate-constraints https://review.openstack.org/556760 | 01:37 |
*** jrist has quit IRC | 03:11 | |
*** jrist has joined #openstack-requirements | 03:12 | |
openstackgerrit | Tony Breeds proposed openstack/requirements master: Add sphinx_feature_classification as a requirement https://review.openstack.org/481294 | 03:27 |
*** udesale has joined #openstack-requirements | 03:29 | |
*** udesale has quit IRC | 03:51 | |
*** hongbin has quit IRC | 04:12 | |
*** cjloader has quit IRC | 04:47 | |
*** cjloader has joined #openstack-requirements | 04:56 | |
*** udesale has joined #openstack-requirements | 05:04 | |
openstackgerrit | Matthew Thode proposed openstack/requirements master: uncap pysaml2 https://review.openstack.org/557434 | 05:15 |
prometheanfire | tonyb: ^ :D | 05:16 |
prometheanfire | hope that works for you | 05:16 |
tonyb | prometheanfire: Thanks | 05:17 |
prometheanfire | tonyb: I'm fine with that time btw, also dirk proposed 2200, he'd have to confirm if 2230 would work | 05:19 |
tonyb | Okay 2230 CEST == 2030 UTC == 0630 AEST == 1530 CDT | 05:24 |
tonyb | If I can calendar right | 05:24 |
prometheanfire | something like that sounds right | 05:25 |
tonyb | prometheanfire: it makes it impossible for coolsvap but he can speak for himself | 05:27 |
*** cjloader has quit IRC | 05:27 | |
prometheanfire | we have 4 fairly diferent timezones to work around, not sure we'll find something perfect | 05:28 |
prometheanfire | I did cc him at least | 05:28 |
tonyb | https://www.worldtimebuddy.com/?qm=1&lid=2147714,4726206,12,1259229&h=2147714&date=2018-4-2&sln=21-22 is when we used to have it, but I'm not a night person and I don't think you're a morning person | 05:30 |
tonyb | so yeah perfect is off the table ;P | 05:30 |
tonyb | if coolsvap opts out then 2030 works for me | 05:30 |
prometheanfire | not quite a morning person | 05:31 |
prometheanfire | ya, we used to do that time | 05:32 |
prometheanfire | anyway, I'm going to bed, long day (had a meeting at 1AM last night :P) | 05:32 |
tonyb | :( | 05:33 |
prometheanfire | ? | 05:34 |
prometheanfire | well, nn | 05:39 |
*** udesale_ has joined #openstack-requirements | 06:13 | |
*** cjloader has joined #openstack-requirements | 06:14 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/requirements master: Updated from generate-constraints https://review.openstack.org/557614 | 06:14 |
dirk | prometheanfire: tonyb : 22:30 local still works for me, although I might occasionally miss it I should be able to join most of the time | 06:15 |
tonyb | dirk: Okay. Cool | 06:16 |
*** udesale has quit IRC | 06:16 | |
*** cjloader has quit IRC | 06:18 | |
*** andreas_s has joined #openstack-requirements | 06:32 | |
*** udesale_ has quit IRC | 06:38 | |
*** udesale has joined #openstack-requirements | 06:39 | |
*** florianf has joined #openstack-requirements | 06:58 | |
*** udesale_ has joined #openstack-requirements | 07:03 | |
*** udesale__ has joined #openstack-requirements | 07:05 | |
*** udesale has quit IRC | 07:07 | |
*** udesale_ has quit IRC | 07:09 | |
*** jrist has quit IRC | 07:27 | |
*** jrist has joined #openstack-requirements | 07:27 | |
*** udesale has joined #openstack-requirements | 07:45 | |
*** udesale__ has quit IRC | 07:47 | |
*** jpich has joined #openstack-requirements | 07:49 | |
*** ralonsoh has joined #openstack-requirements | 07:54 | |
openstackgerrit | Tony Breeds proposed openstack/requirements master: uncap pysaml2 https://review.openstack.org/557434 | 08:04 |
openstackgerrit | Tony Breeds proposed openstack/requirements master: Updated from generate-constraints https://review.openstack.org/557614 | 08:15 |
*** cjloader has joined #openstack-requirements | 09:06 | |
*** cjloader has quit IRC | 09:11 | |
*** cjloader has joined #openstack-requirements | 10:06 | |
*** cjloader has quit IRC | 10:11 | |
*** andreas_s has quit IRC | 11:02 | |
*** andreas_s has joined #openstack-requirements | 11:03 | |
*** andreas_s has quit IRC | 11:07 | |
*** edmondsw has joined #openstack-requirements | 11:30 | |
openstackgerrit | Matthew Edmonds proposed openstack/requirements master: Update pypowervm min to 1.1.13 https://review.openstack.org/557678 | 11:39 |
edmondsw | dhellmann is there a reason that the new per-project lower-constraints.txt files need to include things not in their requirements.txt, or is that just because it's easier to do en masse as you're setting this up? | 11:43 |
*** udesale has quit IRC | 11:46 | |
*** udesale has joined #openstack-requirements | 11:46 | |
*** andreas_s has joined #openstack-requirements | 11:51 | |
*** andreas_s has quit IRC | 11:56 | |
*** udesale_ has joined #openstack-requirements | 12:14 | |
*** udesale has quit IRC | 12:16 | |
*** udesale_ has quit IRC | 12:19 | |
dhellmann | edmondsw : that's how constraints usually work, they list everything. | 12:34 |
edmondsw | dhellmann is something going to maintain that in each project, so it will continue to list everything? | 12:35 |
edmondsw | if so, then that makes sense | 12:35 |
dhellmann | edmondsw : the intent is for project teams to keep that list up to date as they need to adjust their requirements files | 12:38 |
dhellmann | so when you need a new feature of a library, you raise the minimum in say requirements.txt and then also update it in lower-constraints.txt | 12:38 |
edmondsw | IOTW it will rapidly become useless except for things that are listed in requirements.txt... | 12:38 |
dhellmann | well if something in requirements.txt is changed and *needs* a new version of something else, that would need to be raised, too | 12:39 |
edmondsw | I mean, the project-specific lower-constraints is not going to have new things added as they are added in openstack/requirements | 12:40 |
dhellmann | right, we have completely stopped automated updates into projects | 12:40 |
dhellmann | for all dependency lists, not just the new lower-constraints.txt file | 12:40 |
edmondsw | if someone adding a new requirement is going to have to go look at openstack/requirements anyway, why not make them do that for lower-constraints as well as upper? | 12:40 |
edmondsw | and not list unneeded lower-constraints in the meantime | 12:41 |
dhellmann | I'm not sure I understand exactly what you mean | 12:41 |
edmondsw | have project-specific lower-constraints.txt only list things that are in project-specific requirements.txt | 12:42 |
edmondsw | because anything more is unused | 12:42 |
dhellmann | they should only include things that are in requirements.txt *and the dependencies of those things* | 12:42 |
edmondsw | and will likely have to be updated if it ever needs to be used | 12:42 |
dhellmann | I suppose you could try removing the second and third level dependencies and see how that goes, if that's what you mean | 12:42 |
edmondsw | why does it need to list dependencies of those things? | 12:43 |
dhellmann | the point of constraints is to have an exact list of versions to test with. | 12:43 |
dhellmann | if you don't include everything, then a dependency is unconstrained | 12:43 |
edmondsw | k | 12:43 |
dhellmann | which means if I have a requirement for A>=1.0 and A depends on B as B>=1.3 and then B releases 2.0 you might break | 12:44 |
edmondsw | got it, tx | 12:44 |
dhellmann | cool | 12:44 |
dhellmann | we'll probably need to build some tools to help with the lists. I know dirk had something for building the global list that we started with. | 12:45 |
edmondsw | dhellmann maybe I don't have it quite yet... doesn't that example you gave argue for u-c, not l-c? | 12:47 |
dhellmann | both | 12:47 |
dhellmann | if you want to keep the lower-constraints test actually testing at the lower bounds and A 1.0 is not compatible with B 2.0 you have to prevent B 2.0 from accidentally ending up in the lower-constraints test job | 12:47 |
edmondsw | so in that case B might be allowed by u-c, for use with a newer A? | 12:48 |
dhellmann | if A 1.5 is compatible with B 2.0 then in order to add B 2.0 to *upper* constraints you would have to constrain A to 1.5 | 12:48 |
dhellmann | right | 12:48 |
edmondsw | k | 12:48 |
dhellmann | this is all made extra complicated because of the number of dependencies we have and the depth of the graph | 12:49 |
edmondsw | no kidding | 12:49 |
dhellmann | everything would be simpler if we could ensure all distros packaged exactly the version we need | 12:50 |
dhellmann | but we can't, of course, so we're doing the best we can to provide guidance about what is likely to work | 12:50 |
*** odyssey4me has quit IRC | 12:57 | |
*** odyssey4me has joined #openstack-requirements | 12:57 | |
*** edmondsw has quit IRC | 12:58 | |
*** edmondsw has joined #openstack-requirements | 12:59 | |
*** andreas_s has joined #openstack-requirements | 13:01 | |
*** edmondsw has quit IRC | 13:03 | |
*** edmondsw has joined #openstack-requirements | 13:05 | |
*** andreas_s has quit IRC | 13:05 | |
openstackgerrit | Doug Hellmann proposed openstack/requirements master: add validation rules for lower constraints https://review.openstack.org/557034 | 13:37 |
openstackgerrit | Doug Hellmann proposed openstack/requirements master: trigger requirements-check if the lower-constraints.txt file changes https://review.openstack.org/557038 | 13:37 |
openstackgerrit | Doug Hellmann proposed openstack/requirements master: remove debugging print https://review.openstack.org/557338 | 13:37 |
smcginnis | Anyone aware of a subprocess signal.poll() issue that appears to have just recently surfaced? | 13:52 |
*** udesale has joined #openstack-requirements | 13:53 | |
smcginnis | Sorry, select.poll(). Here's a failure - http://logs.openstack.org/73/557673/1/check/build-openstack-sphinx-docs/9bf1f80/job-output.txt.gz#_2018-03-29_12_08_50_350938 | 13:55 |
smcginnis | Oddly, I've only seen it on glance patches so far. | 13:56 |
smcginnis | Guess I'll have to look into what might be done differently there. | 13:56 |
*** udesale has quit IRC | 13:59 | |
prometheanfire | smcginnis: nope, we did update openstackdocstheme (not sphinx) yesterday | 13:59 |
smcginnis | prometheanfire: Hmm, suspicious that this is coming out of there. Thanks, I'll take a quick look at those changes too to see if anything looks suspect. | 14:03 |
*** andreas_s has joined #openstack-requirements | 14:04 | |
prometheanfire | it was a patch bump iirc | 14:07 |
*** andreas_s has quit IRC | 14:10 | |
*** jlvillal has quit IRC | 14:15 | |
openstackgerrit | Matthew Thode proposed openstack/requirements master: uncap pysaml2 https://review.openstack.org/557434 | 14:34 |
prometheanfire | dhellmann: dirk ^ would like your ack on this, want to make sure we DO want to be able to exclude prereleases. | 14:36 |
prometheanfire | oops, had the debug print still in there | 14:37 |
openstackgerrit | Matthew Thode proposed openstack/requirements master: uncap pysaml2 https://review.openstack.org/557434 | 14:37 |
*** cjloader has joined #openstack-requirements | 14:38 | |
*** jlvillal has joined #openstack-requirements | 15:12 | |
*** cjloader has quit IRC | 15:52 | |
openstackgerrit | Merged openstack/requirements master: Add sphinx_feature_classification as a requirement https://review.openstack.org/481294 | 15:55 |
smcginnis | FYI in case anyone else hits it and comes looking, but I've verified the bump to openstackdocstheme 1.20.0 is what has broken glance builds. | 16:30 |
*** florianf has quit IRC | 16:39 | |
*** jpich has quit IRC | 16:46 | |
*** florianf has joined #openstack-requirements | 16:55 | |
openstackgerrit | Doug Hellmann proposed openstack/requirements master: add validation rules for lower constraints https://review.openstack.org/557034 | 17:06 |
openstackgerrit | Doug Hellmann proposed openstack/requirements master: trigger requirements-check if the lower-constraints.txt file changes https://review.openstack.org/557038 | 17:06 |
openstackgerrit | Doug Hellmann proposed openstack/requirements master: remove debugging print https://review.openstack.org/557338 | 17:06 |
*** ralonsoh has quit IRC | 17:19 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/requirements master: update constraint for python-openstackclient to new release 3.15.0 https://review.openstack.org/557782 | 17:37 |
dhellmann | prometheanfire : I don't see an issue excluding pre-releases. Being explicit is probably good, but I would expect them to be excluded by default? | 18:56 |
prometheanfire | they probably are, we don't allow prereleases to be specified in constraints updates or gr >= values, but they could still be pulled in if not using constraints (which is what we want to protect against) | 19:14 |
*** florianf has quit IRC | 19:40 | |
dhellmann | makes sense | 20:10 |
edmondsw | looking for a second +2 on https://review.openstack.org/#/c/557678/ to unblock some commits that need a newer version of pypowervm | 20:13 |
dhellmann | edmondsw : +2a | 20:13 |
edmondsw | dhellmann tx | 20:13 |
openstackgerrit | OpenStack Proposal Bot proposed openstack/requirements master: update constraint for openstackdocstheme to new release 1.20.1 https://review.openstack.org/557821 | 20:16 |
*** edmondsw has quit IRC | 20:47 | |
*** edmondsw has joined #openstack-requirements | 20:47 | |
*** edmondsw has quit IRC | 20:52 | |
lbragstad | prometheanfire: ping | 20:58 |
lbragstad | prometheanfire: do we plan on merging https://review.openstack.org/#/c/557440/1 as is? | 21:03 |
prometheanfire | lbragstad: no | 21:06 |
prometheanfire | lbragstad: that can go away, there will be a bot update with https://review.openstack.org/557434 in it | 21:07 |
lbragstad | oh - neat | 21:07 |
prometheanfire | once that merges we can then move forward on the requirements side raising uc | 21:07 |
prometheanfire | we don't actually need to raise gr, since you passed tests with no code change | 21:07 |
lbragstad | so https://review.openstack.org/#/c/557440/ was merely an exercise to make sure things would pass with pysaml2==4.5.0 | 21:08 |
prometheanfire | yep | 21:08 |
lbragstad | ok - cool | 21:08 |
prometheanfire | you can abaondon if you want | 21:08 |
prometheanfire | or I can | 21:08 |
prometheanfire | whatever | 21:08 |
lbragstad | i have my head in jwt/cryptography stuff yesterday, i'll abandon with a link to this discussion | 21:08 |
prometheanfire | wfm | 21:08 |
prometheanfire | ya, I'm going to start bugging people about moving from pycrypto to cryptography again soon | 21:09 |
lbragstad | and then https://bugs.launchpad.net/keystone/+bug/1750843 will be closed once keystone merges a change with pysaml2 from the opb | 21:09 |
openstack | Launchpad bug 1750843 in OpenStack Global Requirements "pysaml2 version in global requirements must be updated to 4.5.0" [Undecided,New] | 21:09 |
prometheanfire | that bug will be closed when reqs raises at least the UC of pysaml2 | 21:10 |
prometheanfire | at least the reqs portion of it | 21:10 |
prometheanfire | I suspect they'll want us to raise the gr as well (and we probably should) | 21:10 |
prometheanfire | so I guess we'll do that in the followon patch | 21:10 |
lbragstad | ok - so we'll keep the keystone portion of that open for the time being then | 21:11 |
openstackgerrit | Merged openstack/requirements master: Update pypowervm min to 1.1.13 https://review.openstack.org/557678 | 21:45 |
tonyb | prometheanfire: why'd we need to https://review.openstack.org/#/c/557434/6/openstack_requirements/requirement.py@227 in that change? | 22:01 |
*** edmondsw has joined #openstack-requirements | 22:20 | |
*** edmondsw has quit IRC | 22:25 | |
prometheanfire | tonyb: to allow the exclusion of rc1 | 22:27 |
tonyb | prometheanfire: Hmm okay wouldn't it just be better to omit rc1 from the banned list? then everything just ignores it | 22:42 |
tonyb | prometheanfire: Actully I'll do that in a follow-up patch | 22:43 |
*** andreas_s has joined #openstack-requirements | 22:53 | |
prometheanfire | tonyb: we want users to not use rc1 right? | 22:54 |
tonyb | prometheanfire: Correct I *thought* that pip excluded even public prereleases .... which seems to not be the case :/ | 22:55 |
prometheanfire | pip sucks, news at 11 :P | 22:56 |
*** andreas_s has quit IRC | 22:58 | |
openstackgerrit | Tony Breeds proposed openstack/requirements master: Updated from generate-constraints https://review.openstack.org/557614 | 23:20 |
*** andreas_s has joined #openstack-requirements | 23:33 | |
*** andreas_s has quit IRC | 23:38 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!