New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We鈥檒l occasionally send you account related emails.
Already on GitHub? Sign in to your account
New CI #1883
Comments
To start, I removed Travis from the required checks and added the linter, formatter, and Ubuntu test builds. I also turned on "Require branches to be up to date" for all protected branches (it was only on for master). |
Seems a bit better since #1878 was merged, though I'm still seeing some transient failures, e.g. https://github.com/nock/nock/pull/1825/checks?check_run_id=439376631 |
It seems like #1882 is causing builds to fail, e.g. |
It fails because there are no changes to commit. I've added a |
Let me fix that real quick, if anyone has a better idea we can iterate on it |
This is a know issue with GitHub Actions. Any changes made by GitHub Actions (using the Although to be perfectly fair, it did happen in the past that Prettier had unstable formatting, in that case we'd actually have an infinite loop. |
we can use a personal access token from @nockbot for things like that instead.
Can you explain? I'm not sure how this would end up in an infinite loop in this case? After the styling was fixed, there shouldn't be another commit fixing it again |
"unstable" means that Prettier will format its own output in a different way. Here is a list of the past issues of that kind: https://github.com/prettier/prettier/issues?q=is%3Aissue+is%3Aopen+unstable+label%3Aarea%3Aidempotency. As you can see, it happens quite often. If this happens, after formatting and committing, the new execution would format files in a different way, then commit again, then another execution would be triggered, format files in yet another way, ... |
Ohh I see, so it would get back-and-forth, for example? We could add a check on our side to make sure that the format action does not triggered by a commit coming from the same action? |
Yes, we absolutely can! Good idea. I believe we can add an |
Although I'm not sure if that's the right path for |
here is a list of the different events and their payloads https://developer.github.com/v3/activity/events/types/ |
This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. We try to do our best, but nock is maintained by volunteers and there is only so much we can do at a time. Thank you for your contributions. |
@gr2m can this issue be considered complete and closed? |
I think so. We can create new issues for specific problems with the CI, ping me if I can help |
It's cool that we're getting things going with GitHub Actions! Thanks so much to @merlinnot for the legwork (and patience!) 馃弲 and @gr2m for bringing your expertise on this.
I'm trying to get a grasp on the current status of things. I'm seeing some builds which are intermittently failing. For example, on https://github.com/nock/nock/actions/runs/36768497, Node 8 failed on Windows, and on the commit that ran just before https://github.com/nock/nock/actions/runs/36768146, Node 12 failed on Windows. It seems the goal of #1881 is to try to address this, though I'm not sure I understand why an npm upgrade would help in Node 12.
Also, we should update the required checks. Right now Travis is still required, but does not run on pull requests, so that definitely should be removed. Which of the other builders are stable? Should we make them all required except for Windows, to start?
The text was updated successfully, but these errors were encountered: