Repository navigation
Conversation
vilmibm
left a comment
There was a problem hiding this comment.
Thanks for working on this. I appreciate your trying to work around the GitHub platform limitation. I'm concerned about how brittle this is; it's easy to imagine ending up watching the wrong run on a repository with frequent runs.
Is there a way to make this more reliable by filtering runs by workflow file and ref? If not I fear that this PR should be closed for now.
No debate on that here!
That seems like it should be possible, although (being as new as this PR to the code overall) I'm not certain. I'll take a look and report back. |
--watch option to gh workflow run (Fixes #3559)--watch option to gh workflow run
e6cbf42 to
bd2f6cb
Compare
|
@williammartin thanks for the nudge! I think this is ready for another look. In addition to rebasing, I've applied It would still be preferable to close #4001 and resolve this without needing a heuristic, but such an improvement could could of course be made later, under-the-hood of the new |
|
It will fail if there are multiple runs created for same workflow. You might end up watching for other run |
@mik3y: Yeah, having the GitHub REST API endpoints for creating workflow runs return an ID on call is a question / comment / suggestion I've seen within our internal subject matter experts GitHub Actions channel for some time. I've followed up internally to add the GitHub CLI to the list of customers and other users who have requested this. All that said, I'll defer to @williammartin as having been working with you on the issue on your updates and the concern raised about selecting the right workflow run. |
|
@mik3y : I'm going to close this PR rather than keeping this it open indefinitely while I continue to advocate for the GitHub Actions team to enhance Create a workflow dispatch event REST endpoint. There appears to be added complexity with correlating workflow runs depending on whether you use Once again, thank you for your time and patience! ❤️ 🙇 |
|
@andyfeller understood, thanks for pushing on the blocker issue. For anyone coming to this issue/PR in the meantime, fwiw, I've been running a build of this fork as my |
Implementation is somewhat brittle due to #4001, but works adequately.
Fixes #3559. Would be improved by #4001