| -@gerrit:opendev.org- Ian Wienand proposed: [zuul/zuul] web: Project: cleanup query https://review.opendev.org/c/zuul/zuul/+/805742 | 00:29 | |
| @jim:acmegating.com | ianw: each should be a project stanza. 'master' comes from the source context branch name. the project name is omitted which is why it's confusing | 00:30 |
|---|---|---|
| @jim:acmegating.com | ianw: i think the complete information would be something like "openstack/project-config@master#lineA" "openstack/project-config@master#lineB" "airship/airshipctl@master#lineC" | 00:31 |
| @iwienand:matrix.org | hrm ok, thanks | 00:32 |
| @iwienand:matrix.org | i think this also has some issues when looking at a project from another tenant, like diskimage-builder when in zuul tenant. but one thing at a time | 00:33 |
| @iwienand:matrix.org | should be able to fairly easily tidy that page up and convert to modern tabs | 00:33 |
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: [zuul/nodepool] Add commands to export/import image data from ZK https://review.opendev.org/c/zuul/nodepool/+/805743 | 00:52 | |
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: [zuul/nodepool] Support node-attributes in Azure driver https://review.opendev.org/c/zuul/nodepool/+/805707 | 00:52 | |
| -@gerrit:opendev.org- Ian Wienand proposed: [zuul/zuul] Add source context to project API return https://review.opendev.org/c/zuul/zuul/+/805746 | 01:29 | |
| @iwienand:matrix.org | > <@jim:acmegating.com> ianw: each should be a project stanza. 'master' comes from the source context branch name. the project name is omitted which is why it's confusing | 01:33 |
| https://review.opendev.org/c/zuul/zuul/+/805746 should help to disambiguate | ||
| -@gerrit:opendev.org- Ian Wienand proposed: [zuul/zuul] Add source context to project API return https://review.opendev.org/c/zuul/zuul/+/805746 | 03:12 | |
| -@gerrit:opendev.org- Simon Westphahl proposed on behalf of Felix Edel: [zuul/zuul] Make reporting asynchronous https://review.opendev.org/c/zuul/zuul/+/691253 | 05:21 | |
| @avass:vassast.org | corvus: do we wanna start deploying the zuul-operator docs so it's easier to find, or should that wait until a v1.0.0? | 06:44 |
| -@gerrit:opendev.org- Simon Westphahl proposed on behalf of Felix Edel: [zuul/zuul] Make reporting asynchronous https://review.opendev.org/c/zuul/zuul/+/691253 | 08:38 | |
| -@gerrit:opendev.org- Zuul merged on behalf of James E. Blair https://matrix.to/#/@jim:acmegating.com: [zuul/zuul] Adjust test_autohold to catch stats errors https://review.opendev.org/c/zuul/zuul/+/805416 | 11:24 | |
| -@gerrit:opendev.org- Zuul merged on behalf of Pierre-Louis Bonicoli: [zuul/zuul] Escape raw percent in 'dburi' value https://review.opendev.org/c/zuul/zuul/+/797660 | 11:44 | |
| -@gerrit:opendev.org- Simon Westphahl proposed: | 12:05 | |
| - [zuul/zuul] Implement ABC for caching changes in Zookeeper https://review.opendev.org/c/zuul/zuul/+/805835 | ||
| - [zuul/zuul] wip: Add source interfaces for cached changes https://review.opendev.org/c/zuul/zuul/+/805836 | ||
| - [zuul/zuul] wip: Gerrit change cache in Zookeeper https://review.opendev.org/c/zuul/zuul/+/805837 | ||
| - [zuul/zuul] wip: Github change cache in Zookeeper https://review.opendev.org/c/zuul/zuul/+/805838 | ||
| - [zuul/zuul] dnm: make tests pass temporarily https://review.opendev.org/c/zuul/zuul/+/805839 | ||
| -@gerrit:opendev.org- Simon Westphahl proposed: | 12:58 | |
| - [zuul/zuul] Implement ABC for caching changes in Zookeeper https://review.opendev.org/c/zuul/zuul/+/805835 | ||
| - [zuul/zuul] wip: Add source interfaces for cached changes https://review.opendev.org/c/zuul/zuul/+/805836 | ||
| - [zuul/zuul] wip: Gerrit change cache in Zookeeper https://review.opendev.org/c/zuul/zuul/+/805837 | ||
| - [zuul/zuul] wip: Github change cache in Zookeeper https://review.opendev.org/c/zuul/zuul/+/805838 | ||
| - [zuul/zuul] dnm: make tests pass temporarily https://review.opendev.org/c/zuul/zuul/+/805839 | ||
| - [zuul/zuul] wip: Resolve change dependency references lazily https://review.opendev.org/c/zuul/zuul/+/805844 | ||
| -@gerrit:opendev.org- Simon Westphahl proposed: | 13:04 | |
| - [zuul/zuul] wip: Add source interfaces for cached changes https://review.opendev.org/c/zuul/zuul/+/805836 | ||
| - [zuul/zuul] wip: Resolve change dependency references lazily https://review.opendev.org/c/zuul/zuul/+/805844 | ||
| - [zuul/zuul] wip: Gerrit change cache in Zookeeper https://review.opendev.org/c/zuul/zuul/+/805837 | ||
| - [zuul/zuul] wip: Github change cache in Zookeeper https://review.opendev.org/c/zuul/zuul/+/805838 | ||
| - [zuul/zuul] dnm: make tests pass temporarily https://review.opendev.org/c/zuul/zuul/+/805839 | ||
| @apevec:matrix.org | tristanC: ++ matrix-gerritbot is very nice! | 13:39 |
| @jim:acmegating.com | > <@avass:vassast.org> corvus: do we wanna start deploying the zuul-operator docs so it's easier to find, or should that wait until a v1.0.0? | 14:31 |
| i think we're pretty close to a 1.0 and we should publish the docs | ||
| @clarkb:matrix.org | @corvus | 16:54 |
| @clarkb:matrix.org | er thats an annoying UX thing. I wanted to select your user when it popped up but return just sent the message. | 16:55 |
| @clarkb:matrix.org | corvus: did you see the need to rebase https://review.opendev.org/c/zuul/zuul/+/800854/ and children? https://review.opendev.org/c/opendev/system-config/+/803992 is the opendev side change to accommodate those updates too | 16:55 |
| @jim:acmegating.com | Clark: at least you can edit your msg :) | 16:57 |
| @jim:acmegating.com | Clark: thanks will rebase | 16:58 |
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: | 17:07 | |
| - [zuul/zuul] Add commands to export/import keys to/from ZK https://review.opendev.org/c/zuul/zuul/+/800854 | ||
| - [zuul/zuul] Add copy-keys and delete-keys zuul client commands https://review.opendev.org/c/zuul/zuul/+/803973 | ||
| - [zuul/zuul] Add delete-state command to delete everything from ZK https://review.opendev.org/c/zuul/zuul/+/804304 | ||
| @clarkb:matrix.org | Looks like the first change needs updated reviews. I'll do that after this meeting | 17:13 |
| @jim:acmegating.com | only changes are in test helper methods | 17:17 |
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: [zuul/nodepool] Add commands to export/import image data from ZK https://review.opendev.org/c/zuul/nodepool/+/805743 | 17:28 | |
| @jim:acmegating.com | apparently zuul jobs are failing because of "can't find Rust compiler" https://zuul.opendev.org/t/zuul/build/a78666ca0bcc4a39bd087ea3c9c3eed4 | 17:57 |
| @jim:acmegating.com | is this a wheel upload out of order issue? | 17:58 |
| @avass:vassast.org | corvus: yep | 18:03 |
| @avass:vassast.org | As far as I remember at least | 18:03 |
| @clarkb:matrix.org | cryptography probably as that requires rust now | 18:06 |
| @clarkb:matrix.org | but seems that one wasn't released recently | 18:07 |
| @jim:acmegating.com | oh, then hrm, maybe this is some new issue? | 18:07 |
| @clarkb:matrix.org | oh wait no someone is publishing cryptography318 as a package on pypi and that is the first result now | 18:08 |
| @clarkb:matrix.org | actual cryptography did release about an hour ago says pypi | 18:08 |
| @jim:acmegating.com | okay... so the procedure here is to make a sandwich, get some coffee, do the dishes, then recheck? | 18:09 |
| @clarkb:matrix.org | https://pypi.org/project/cryptography/#files you can see the tarball is older than the sdists | 18:09 |
| @clarkb:matrix.org | I think if you recheck now you will be good | 18:09 |
| @jim:acmegating.com | if i'm reading those human timestamps right, there's a ~21 minute window where this can happen? | 18:10 |
| @jim:acmegating.com | ah if i read the actual timestamps, looks more like 35 minutes | 18:11 |
| @fungicide:matrix.org | yep, it's a semi-common race between cryptography sdist and wheel publication. if you try to install it on a platform for which there's not yet a wheel available, you'll need rust in order to compile the relevant bindings | 18:30 |
| @fungicide:matrix.org | how long will vary by platform, if there are longer delays publishing wheels for some platforms than others | 18:31 |
| @jim:acmegating.com | is it possible for them to reverse the order of publication? | 18:32 |
| @jim:acmegating.com | (and would that fix this issue?) | 18:32 |
| @fungicide:matrix.org | probably, but they'd need some dependency mechanism to hold back sdist publication until their various wheel builders have finished and the results have been uploaded | 18:33 |
| @fungicide:matrix.org | also that *might* be seen as distributing compiled binaries withiout accompanying source (though that's a stretch, since they have a public git repo) | 18:33 |
| @jim:acmegating.com | maybe we should install rust compilers? | 18:34 |
| @fungicide:matrix.org | it would be technically correct, and helpful for people who try to install zuul on platforms where there is no available cryptography wheel (e.g. when trying out beta releases of python) | 18:34 |
| @avass:vassast.org | There's already an ensure-rust role so it should be pretty easy | 18:35 |
| @fungicide:matrix.org | but yeah, it's mainly a question of whether it's a case worth optimizing for | 18:36 |
| @fungicide:matrix.org | and whether it makes sense to install rust all the time but only use it for a few minutes out of the year | 18:37 |
| @avass:vassast.org | I don't think it takes very long to install so it may be worth it. could also be possible to check if the wheel is older than the sdist and only install rust if it is | 18:38 |
| @jim:acmegating.com | or stick it on the image | 18:39 |
| @jim:acmegating.com | but, ideally, if we can install rustc quickly from the mirrors, that would be nice | 18:39 |
| @fungicide:matrix.org | a big part of why there's not been much done to address this pain point in the python packaging ecosystem in recent years is a prevailnig shift to assume most developers are going to utilize javascript/go style "lock files" instead of open-ended dependencies. there's a pep in progress now for encoding the lock file idea more officially into package metadata as well | 18:42 |
| @jim:acmegating.com | that's not really a solution, it just shifts the problem. you won't be able to run the lock process during the 35 minute window. | 18:43 |
| @jim:acmegating.com | but it does hide the issue from developers who are unconcerned with keeping their cryptographic libraries up to date. 🙄 | 18:44 |
| @fungicide:matrix.org | right, i think the assumption is that project maintainers will only increase the version they're locking once they know the relevant release artifacts are in place (which you're ruight is a continued assumption about a lack of process automation) | 18:45 |
| @fungicide:matrix.org | one workaround might be to override pip install with --prefer-binary so that it will use older cryptography if there's no available wheel | 18:47 |
| @fungicide:matrix.org | there's also an --only-binary option which accepts some parameters allowing you to control it for specific dependencies | 18:48 |
| @fungicide:matrix.org | so may be able to just do `pip install --only-binary=cryptography .` | 18:49 |
| @clarkb:matrix.org | we use prefer-binary on the arm64 nodepool builds because the cost of using an sdist is so high there | 18:51 |
| @clarkb:matrix.org | corvus: is the zk management stack the one that hit the cryptography problem? | 20:03 |
| @iwienand:matrix.org | i have offered to build and publish cryptography's sdists and wheels | 20:03 |
| @iwienand:matrix.org | windows and osx make it a pita | 20:04 |
| @jim:acmegating.com | Clark: yes | 20:05 |
| @jim:acmegating.com | Clark: the change i rechecked passed; i think it's safe for you to review | 20:06 |
| @jim:acmegating.com | if we want to carry over tobiash's review we could probably approve it | 20:07 |
| @iwienand:matrix.org | corvus: https://review.opendev.org/c/zuul/zuul/+/805746 is the one from the source_context yesterday, it had to use toDict() on the context instead of pass the object directly | 20:11 |
| @jim:acmegating.com | yay testing | 20:11 |
| @iwienand:matrix.org | i also responded on https://review.opendev.org/c/zuul/zuul/+/804956 which is the treeview change; there's an example of an expand button in https://review.opendev.org/c/zuul/zuul/+/805340/1 | 20:12 |
| @iwienand:matrix.org | but i did learn something about the "X" in the PF search box with https://review.opendev.org/c/zuul/zuul/+/805581 which implements a filter for projects; it only appears if you set a value via onChange event in the search box | 20:13 |
| @iwienand:matrix.org | i think it might be worth raising with patternfly, as that can cause render() to re-run constantly. it seems a bit much | 20:14 |
| @clarkb:matrix.org | > <@jim:acmegating.com> Clark: the change i rechecked passed; i think it's safe for you to review | 20:52 |
| Yup I think the whole stack has my +2 now | ||
| @clarkb:matrix.org | I'm noticing mhuins web login stack seems to conflict with a number of other web ui changes. I haven't followed that work too closely. Should we consider picking a window to land that and then we can rebase and resolve conflicts and everything web after that has to rebase to resolve conflicts? | 21:54 |
| @clarkb:matrix.org | ianw's https://review.opendev.org/c/zuul/zuul/+/805746 is super tiny but conflicts according to gerrit as an example | 21:55 |
| -@gerrit:opendev.org- Zuul merged on behalf of James E. Blair https://matrix.to/#/@jim:acmegating.com: [zuul/zuul] Add commands to export/import keys to/from ZK https://review.opendev.org/c/zuul/zuul/+/800854 | 22:17 | |
| -@gerrit:opendev.org- Zuul merged on behalf of James E. Blair https://matrix.to/#/@jim:acmegating.com: | 22:21 | |
| - [zuul/zuul] Add copy-keys and delete-keys zuul client commands https://review.opendev.org/c/zuul/zuul/+/803973 | ||
| - [zuul/zuul] Add delete-state command to delete everything from ZK https://review.opendev.org/c/zuul/zuul/+/804304 | ||
| -@gerrit:opendev.org- Ian Wienand proposed: [zuul/zuul-jobs] build-docker-image: Add flag to use BuildKit https://review.opendev.org/c/zuul/zuul-jobs/+/805911 | 22:31 | |
| @clarkb:matrix.org | ianw: https://review.opendev.org/c/zuul/zuul/+/804602 looks much better to me now. Thanks | 22:35 |
| @jim:acmegating.com | Clark, ianw: that looks like it moves everything pretty far to the left and expands the space between the items again? | 22:56 |
| @iwienand:matrix.org | ahh sorry yes that has dropped the css override | 23:00 |
| @clarkb:matrix.org | hrm it didnt' seem as bad as before but that could also just be my eyeballs not remembering | 23:05 |
| @clarkb:matrix.org | ianw: https://review.opendev.org/c/zuul/zuul/+/805581 has some notes on it. Sorry my js foo isn't great so it is a little handwavy | 23:05 |
| @iwienand:matrix.org | thanks i'll have a look | 23:07 |
| -@gerrit:opendev.org- Ian Wienand proposed: [zuul/zuul] web: JobVariant : pull description into a card https://review.opendev.org/c/zuul/zuul/+/804602 | 23:08 | |
| @iwienand:matrix.org | > <@gerrit:opendev.org> Ian Wienand proposed: [zuul/zuul] web: JobVariant : pull description into a card https://review.opendev.org/c/zuul/zuul/+/804602 | 23:08 |
| this drops the xl margin around the entire tab view, and just puts it on the description list. i think that's more consistent. and it should restore the line height thing | ||
| @clarkb:matrix.org | cool I'll check it in the site preview once that is built | 23:09 |
| @iwienand:matrix.org | thanks i hadn't noticed the search pulldown thing. that's a bit annoying, i'm not sure how we make the height long enough to always show that | 23:11 |
| @iwienand:matrix.org | i guess a min-height on the table? | 23:11 |
| @clarkb:matrix.org | or can you split the search object out of the section with the results and give it its own section ? maybe that will push the result rows down when you open the dropdown | 23:12 |
| @iwienand:matrix.org | for the enter on the search; i'm not 100% sure ... not sure i know how to simulate the click | 23:13 |
| @clarkb:matrix.org | it isn't critical and the existing page doesn't have any search so still an improvement. But noticed that when testing it and thought it would be a nice update if possible | 23:24 |
| -@gerrit:opendev.org- Ian Wienand proposed: [zuul/zuul-jobs] build-docker-image: Add flag to use BuildKit https://review.opendev.org/c/zuul/zuul-jobs/+/805911 | 23:51 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!