You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When a build fails, GitHub Actions defaults to sending a notification (and/or email) to the last committer to that repository. So, for example, I'm the last committer to CMILES, so I get the emails for the nightly build failures (known issue openforcefield/cmiles#39). This criterion is inconsistent, and instead we should configure each repo to have a single (or few) maintainer(s) who get these notifications and are responsible for investigating.
I'm actually not sure if this is configurable technically, the best I could find in the docs doesn't satisfy this. I recalling seeing "email this person" type of content in a CI YAML before, but maybe that was Travis.
I know we've spent a lot of time configuring Actions lately, but from my perspective, this is the only thing missing for a unified structure across the organization.
The text was updated successfully, but these errors were encountered:
Yep, I dislike this too. I would like to have something that at leases does something like using the code_owners file to choose who to email (as mentioned in this thread). This is specially useful with cron jobs. Maybe a Slack channel could be set up for the cron job notifications?
When a build fails, GitHub Actions defaults to sending a notification (and/or email) to the last committer to that repository. So, for example, I'm the last committer to
CMILES
, so I get the emails for the nightly build failures (known issue openforcefield/cmiles#39). This criterion is inconsistent, and instead we should configure each repo to have a single (or few) maintainer(s) who get these notifications and are responsible for investigating.I'm actually not sure if this is configurable technically, the best I could find in the docs doesn't satisfy this. I recalling seeing "email this person" type of content in a CI YAML before, but maybe that was Travis.
I know we've spent a lot of time configuring Actions lately, but from my perspective, this is the only thing missing for a unified structure across the organization.
The text was updated successfully, but these errors were encountered: