| frickler | not sure where to best ask this, but how is the openstack/openstack repo maintained? I was crosschecking the xstatic-font-awesome retirement and see it still being listed as submodule there | 08:53 |
|---|---|---|
| frickler | unrelated: the list of runtimes at the bottom of https://governance.openstack.org/tc/reference/project-testing-interface.html could also use better sorting | 10:07 |
| *** gthiemon1e is now known as gthiemonge | 12:24 | |
| gmann | frickler: openstack/openstack should be updated on those repo additional/removal | 12:49 |
| frickler | gmann: "should be" ... done by someone or by some job which might be broken? | 12:50 |
| gmann | frickler: it was job but I need to check that | 12:52 |
| fungi | yeah, i poked around briefly trying to find the update job, but didn't turn up where it's defined | 12:57 |
| frickler | I found https://opendev.org/openstack/openstack/src/branch/master/generate-gitmodules.py but no reference to that in codesearch | 13:02 |
| knikolla | o/ | 13:05 |
| gmann | but does it remove things? I can see tempest-lib also there which was retired a year ago | 13:06 |
| knikolla | gouthamr: IIRC, we don't usually record PTG sessions. But there have been some cases. | 13:07 |
| fungi | gmann: frickler: i doubt it removes retired projects since they're still openstack project governed repositories, even if they're not expected to merge new changes very often (they're not entirely read-only, the tc still has access to alter them if necessary) | 13:08 |
| frickler | fungi: xstatic-font-awesome no longer is in https://opendev.org/openstack/governance/raw/reference/projects.yaml and if I read https://opendev.org/openstack/openstack/src/branch/master/generate-gitmodules.py#L54 correctly that should be a "git rm" | 13:15 |
| * frickler now tries to run that script locally with only echoing those commands | 13:16 | |
| frickler | or maybe in some hours, actually cloning that repo doesn't seem to be fast | 13:17 |
| gmann | I do not think it should keep retired one as they are no longer 'component of openstack ', 'not tested by zuul' | 13:17 |
| gmann | I was wrong that it does remove things also which i feel it should | 13:18 |
| fungi | frickler: gmann: a git log of the .gitmodules file indicates it was last updated manually with https://review.opendev.org/736012 | 13:19 |
| fungi | that could be why we're not finding any automation... there isn't any? | 13:19 |
| frickler | oh, yes, if that has been a manual ajaeger-task before, that would explain it | 13:20 |
| fungi | maybe hberaud remembers reviewing those up to a few years ago | 13:20 |
| frickler | running manually lists roughly 50 additions and 20 removals pending | 13:21 |
| gmann | ah, right | 13:21 |
| fungi | we could probably create a proposal job triggered from merges to governance that alter relevant files (maybe just projects.yaml) | 13:21 |
| frickler | hmm, I tried to create an update manually, but even "git review -s" fails with "Description: fatal: Upload denied for project 'openstack/openstack'" | 13:56 |
| fungi | i think there's an acl limiting who can propose changes | 13:58 |
| gmann | yeah https://github.com/openstack/project-config/blob/master/gerrit/acls/openstack/openstack.config | 13:58 |
| frickler | but why does that even block -s? | 13:59 |
| gmann | we can add TC group too as it is repo under TC to maintain? | 13:59 |
| frickler | from the acl that could be a release team task, guess I'll ask over there first | 14:01 |
| fungi | frickler: git review -s performs a test push to confirm you have sufficient rights to propose changes | 14:04 |
| fungi | so i guess it's working as intended, since it errored for that reason | 14:04 |
| frickler | oh, I wasn't aware of that. but likely makes sense to fail early, rather than after having done all the work to prepare a patch | 14:06 |
| fungi | the most common place new users run into it is when they haven't agreed to the openinfra icla and try to push changes for a cla-enforced project | 14:11 |
| fungi | so it helps that the first recommended step in starting to work on a change in a new repo is git review -s, since it catches that straight away | 14:16 |
| clarkb | do we know if anyone is getting value out of that repo? | 14:22 |
| clarkb | the original idea behind it was to have a linear historical record of changes landing across openstack projects | 14:22 |
| clarkb | in particular this was envisioned as being useful for continuous deployment systems needing to bisect issues. | 14:23 |
| knikolla | interesting, i didn't know it existed | 14:26 |
| fungi | clarkb: apparently the cncf "landscape" gets value out of the github replica of it, since cncf/lf assumes open source projects are all monorepos and exist on github | 14:32 |
| fungi | so to have openstack be included in the landscape it had to be added as a single git repository there | 14:33 |
| PagliaccisCloud | #kata now discussing runtime-rs | 14:34 |
| PagliaccisCloud | Yikes wrong channel - sorry! | 14:34 |
| fungi | PagliaccisCloud: you want #openinfra-events | 14:34 |
| clarkb | fungi: interesting. I'm pretty sure openstack isn't unique in having more than one repo. I wonder what others do about that | 14:35 |
| fungi | in some cases they just pick one repository, in some cases they're listed as distinct projects | 14:37 |
| opendevreview | Vishal Manchanda proposed openstack/governance master: Add xstatic-angular-fileupload as Horizon team deliverables https://review.opendev.org/c/openstack/governance/+/873845 | 17:25 |
| opendevreview | James Slagle proposed openstack/governance master: TripleO: switch to distributed project leadership https://review.opendev.org/c/openstack/governance/+/878799 | 21:41 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!