| *** haleyb|out is now known as haleyb | 13:55 | |
| opendevreview | Nicolas Hicher proposed openstack/openstack-zuul-jobs master: openafs-rpm-package-build: workaround for dkms.conf option https://review.opendev.org/c/openstack/openstack-zuul-jobs/+/951527 | 14:12 |
|---|---|---|
| dansmith | remind me, we're not supposed to create new x-openstack repos right? | 14:17 |
| fungi | what's an x-openstack repo? | 14:20 |
| fungi | you mean add repositories in the x/ git namespace? | 14:21 |
| dansmith | yeah | 14:21 |
| fungi | if so, no, it would b better to pick a more descriptive namespace for whatever you're creating if it's not suitable for inclusion in openstack/ | 14:21 |
| fungi | x/ is where we stuck all th projects whose maintainers never responded when we asked what namespaces they wanted to be moved into when we cleared unofficial repos out of openstack/ years ago | 14:22 |
| dansmith | the docs kinda imply it's also appropriate for catch-all random stuff | 14:22 |
| dansmith | "or use the catch-all x namespace if you cannot decide." | 14:22 |
| dansmith | anyway, | 14:22 |
| fungi | there's really no additional work to use a different namespace, but yeah we may have some old references that should be cleaned up in docs, if you've found some | 14:23 |
| dansmith | I know I just did this with os-test images in the last year, but remind me: it requires TC approval to create right? | 14:23 |
| dansmith | ...in the openstack namespace I guess | 14:23 |
| fungi | right | 14:24 |
| fungi | though if there's a ptl approving addition into a maintained project team or a sig chair adding it to a sig then it's essentially rubber-stamped by the tc | 14:24 |
| dansmith | is it more expedient if it's another namespace? | 14:24 |
| fungi | short turnaround house ruls | 14:24 |
| fungi | if it's another (new) namespace then there's no approval needed | 14:25 |
| fungi | if it's another existing namspace, then some coordination with other people using that namspace would be a good idea | 14:25 |
| dansmith | I just want a repo I can run unit tests with to hold a script that operators can use to avoid it being in nova proper | 14:26 |
| dansmith | as much as I hate github, perhaps it's just best to go that route, it would just be nice to keep it in the family | 14:27 |
| opendevreview | Nicolas Hicher proposed openstack/openstack-zuul-jobs master: openafs-rpm-package-build: workaround for dkms.conf option https://review.opendev.org/c/openstack/openstack-zuul-jobs/+/951527 | 14:27 |
| fungi | i guess openstack/nova doesn't have a contrib directory or anything | 14:27 |
| dansmith | it does, and it's there now | 14:27 |
| fungi | in some of our projects we stick stuff like that in tools/ and don't include it into any built packages | 14:28 |
| dansmith | I'm trying to get it out of there so I can iterate on it a bunch and run some tests without generating a ton of git traffic and needs-core-review for an unsupported thing :) | 14:28 |
| fungi | self-approval exceptions for things in that directory could be another route? | 14:28 |
| fungi | just seems like splitting it into another repo to avoid the same reviwe policies is overenginering a technical solution to a human challenge | 14:29 |
| dansmith | yeah, I'm just trying to extract it from the project as much as possible... for more than just the review reasons, etc, but fair enough | 14:32 |
| opendevreview | Dan Smith proposed openstack/project-config master: Add nova-contrib https://review.opendev.org/c/openstack/project-config/+/951570 | 16:12 |
| dansmith | I don't understand why this works for nova-specs but not nova-contrib ... https://zuul.opendev.org/t/openstack/build/8f46f4669d4443d49428d93424e813c9 | 17:23 |
| fungi | dansmith: nova-specs has its own acl: https://opendev.org/openstack/project-config/src/branch/master/gerrit/acls/openstack/nova-specs.config | 17:32 |
| clarkb | gmaan left a comment about it I Think | 17:33 |
| JayF | ++ the comment on that change reflects what I suspected as well | 17:33 |
| gmaan | dansmith: yeah it is missing acl file https://review.opendev.org/c/openstack/project-config/+/951570/1/gerrit/projects.yaml#4305 | 17:34 |
| dansmith | oh, I assumed this worked differently, but okay makes sense | 17:34 |
| dansmith | sorry missed gmaan's comment, obviously :D | 17:35 |
| dansmith | okay so probably best thing is to just override to nova's acl instead of duplicating | 17:36 |
| dansmith | that's what I assumed the default was, | 17:36 |
| fungi | agred | 17:36 |
| dansmith | although I have no explanation for how that would have worked except for ... in my head | 17:36 |
| fungi | e | 17:36 |
| opendevreview | Dan Smith proposed openstack/project-config master: Add nova-contrib https://review.opendev.org/c/openstack/project-config/+/951570 | 17:44 |
| *** haleyb_ is now known as haleyb | 19:37 | |
| opendevreview | Merged openstack/openstack-zuul-jobs master: openafs-rpm-package-build: workaround for dkms.conf option https://review.opendev.org/c/openstack/openstack-zuul-jobs/+/951527 | 21:34 |
Generated by irclog2html.py 4.0.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!