-
Notifications
You must be signed in to change notification settings - Fork 64.9k
Description
What article on docs.github.com is affected?
https://docs.github.com/en/free-pro-team@latest/actions/learn-github-actions/managing-complex-workflows#creating-dependent-jobs and/or https://docs.github.com/en/free-pro-team@latest/actions/reference/workflow-syntax-for-github-actions#jobsjob_idneeds
What part(s) of the article would you like to see updated?
In https://docs.github.com/en/free-pro-team@latest/actions/learn-github-actions/managing-complex-workflows#creating-dependent-jobs, the possibility of constraining jobs to run sequentially without the failure of earlier jobs in the sequence preventing the later jobs from running: If one of the jobs fails, all dependent jobs are skipped; however, if you need the jobs to continue, you can define this using the if conditional statement.
I had expected, thus, to see an example of how to do this, either directly here or in one of the two linked-to articles (either https://docs.github.com/en/free-pro-team@latest/actions/reference/workflow-syntax-for-github-actions#jobsjob_idif or https://docs.github.com/en/free-pro-team@latest/actions/reference/workflow-syntax-for-github-actions#jobsjob_idneeds).
It is not clear what sort of "if" statement is called for, nor whether it should supplement or replace the "needs" directive. The https://docs.github.com/en/free-pro-team@latest/actions/reference/workflow-syntax-for-github-actions#jobsjob_idneeds article reiterates the possibility of constraining jobs to run in sequence without failing jobs preventing later jobs from running: If a job fails, all jobs that need it are skipped unless the jobs use a conditional statement that causes the job to continue. This at least suggests that the "if" statement supplements the "needs" directive, but again, there is no example given of how to do this.
Since the Creating dependent jobs article occurs in the Learn GitHub Actions section of the documentation, it would probably be better to just omit the comment about allowing dependent jobs to run when their precursors fail, or at least to provide or link to an example the clarifies how to set this up.
Additional information
Content Plan
Linked here