SamThursfield[m] | true, that would also allow simplifying the Github Actions stuff for BuildStream, as the BuildBox CI could update the buildstream-buildbox PyPI package directly | 08:53 |
---|---|---|
SamThursfield[m] | would require quite a few days to rework and retest it all though | 08:53 |
SamThursfield[m] | FWIW, the docker image seems to be working. Tested in buildstream.git by running cd doc/examples/autotools; ../../../contrib/bst-here -p build | 13:41 |
SamThursfield[m] | same with -j nightly | 13:42 |
SamThursfield[m] | while looking at the docker-images repo, fedora 36 needs updating to fedora 38 | 13:47 |
SamThursfield[m] | pushed https://gitlab.com/BuildStream/buildstream-docker-images/-/merge_requests/218 to hopefully solve those things | 14:00 |
SamThursfield[m] | oh, https://gitlab.com/BuildStream/buildstream-docker-images/-/merge_requests/217 already exists to fix toolbox issues | 14:15 |
SamThursfield[m] | will rerun the CI for that and merge | 14:15 |
SamThursfield[m] | seems CI is failing in the "release" jobs, I guess because of missing credential ? | 15:47 |
SamThursfield[m] | i think it should be ok to just merge, but i don't have permission to do that anyway | 15:48 |
SamThursfield[m] | jjardon: juergbi does one of you have permission to land https://gitlab.com/BuildStream/buildstream-docker-images/-/merge_requests/217 without the 'release' CI jobs passing ? | 15:48 |
SamThursfield[m] | i assume the failure is because the CI ran in fork ddrake/buildstream-docker-images and not in the upstream project - | 15:49 |
SamThursfield[m] | or, i've included it in https://gitlab.com/BuildStream/buildstream-docker-images/-/merge_requests/218 so if CI passes there we could land it as part of my MR | 15:54 |
SamThursfield[m] | ah yeah - CI probably failing on dsd's PR because the branch name is 'master' so the CI enables extra jobs to do a release | 15:54 |
juergbi | SamThursfield[m]: it doesn't seem like I can force merge this. maybe it would be necessary to temporarily change the project setting that requires CI to succeed | 16:36 |
nanonyme | Sam Thursfield: I suggest closing the MR, stop using master branch name and make a new MR | 18:05 |
nanonyme | The only reason you have failing pipeline is you use master branch name | 18:06 |
nanonyme | juergbi: no reason to force or change project settings. It's a user error. | 18:10 |
nanonyme | Well, to be honest it would be great to limit it to master branch only of the actual repo but I don't know if Gitlab supports that | 18:11 |
nanonyme | In any case, use feature branches with names | 18:11 |
nanonyme | Actually it might. Something like | 18:13 |
nanonyme | ``` | 18:13 |
nanonyme | rules: | 18:13 |
nanonyme | - if: '$CI_PROJECT_NAMESPACE == "BuildStream" && $CI_COMMIT_BRANCH == $CI_DEFAULT_BRANCH' | 18:13 |
nanonyme | ``` | 18:13 |
nanonyme | https://gitlab.com/BuildStream/buildstream-docker-images/-/merge_requests/219 | 20:30 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!