jajik commented on code in PR #272: URL: https://github.com/apache/commons-daemon/pull/272#discussion_r2269741383
########## .github/workflows/codeql-analysis-cpp.yml: ########## @@ -17,7 +17,6 @@ name: "CodeQL CPP" on: push: - branches: [ master ] Review Comment: > An example illustrates the issue best: take a look at the workflow runs for each commit in [apache/commons-compress#685](https://github.com/apache/commons-compress/pull/685). Every time a commit was pushed to address a review comment, the workflows ran **twice**, once for the `push` event and again for the `pull_request` event, doubling the CI load unnecessarily. I understand what you mean, but 1. The CI load in the project is "doubled" because of using branches (in the project) instead of forks. 2. I don't see what's the issue/downside with the "duplication" in case of using branches. 3. Using branches is restricted to maintainers/committers only (?), and yet most of the PRs are from forks anyway. (To be honest, dependabot has a lot of PRs too, but we can ignore its branches for the push trigger.) > As I mentioned above, I understand the frustration of testing PRs that modify workflows: it’s challenging both to prepare and to review, since the effect is **not** visible until after the PR is merged. However, such workflow-changing PRs are the exception, not the norm. But the issue at hand is not limited to workflow changing PRs, quite contrary. It's about testing of **all** changes **before** a PR is created and its workflow is approved to run in this repository -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: issues-unsubscr...@commons.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
