Closed obnoxxx closed 1 year ago
@phlogistonjohn , @synarete : an attempt at adding a commitlint ci check. this check fails in a strange way though ...
commitlint complains about errors in the commitlintrc.json. I don't reaaly understand what's wrong ...
the jso does in fact seem to be wrongly formatted ( in the list of components in fact. strange ... I don't seem to know enough about json to fix it ... 🤷
python3 -m json.tool < .commitlintrc.json
Expecting value: line 26 column 2 (char 411)
@raghavendra-talur fyi 😄
Managed to fix the json! 😄 Thanks for the hint, @raghavendra-talur
@spuiuk , @synarete , @anoopcs9 , @phlogistonjohn : for review, it is most important to check and agree to the list of components encoded in the commitlintrc.json file ...
@obnoxxx & @synarete I'd like to kindly request that we focus our discussion on one commitlint PR - this one in particular since it already has the most comments. Once we've discussed this PR sufficiently I think many of the points of discussion should then "flow down" to the similar PRs on the other SINK projects.
ToDos from comments and direct discussions:
@phlogistonjohn , @synarete - fyi, a summary of our discussion ...
fwiw, PR #274 has been created to propose the alternative using gitlint ...
I added a test commit to see commitlint fail. here we go:
You have commit messages with errors
⧗ input: foobar: introduce typo
test commit to hopefully trigger a check error in PR #272 ...
Signed-off-by: Michael Adam obnox@samba.org ✖ type must be one of [docs, resources, planner, config, ci, api, github, controllers, tests, test utils, tests/integration, ?> > i ntegration tests, hack, main, dockerfile, makefile] [type-enum]
Looks quite reasonabe;, eexcept for: why does it list [type-enum]
here? and why makefile]
instead of makefile
?...
this is not needed any more since we enabled gitlint by merging PR #298
The commitlint config and workflow definition is inspired by and essentially copied from the rook one.
Signed-off-by: Michael Adam obnox@samba.org