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
I have no idea of how hard is it implement, but I've found the following:
When I propose to add some libraries to the collection, the whole collection is rechecked. So, when any library in the collection fails to build, the whole result of CI check for the PR is considered to be bad.
What I propose is to try to narrow the set of checked packages to changed/added packages and those packages that are anyhow influenced by those changes ones. This would lead to more adequate results of CI tests upon PRs.
The text was updated successfully, but these errors were encountered:
I'll have to think about this. Should be possible to in pack-admin compare HEAD.toml of the PR against HEAD.toml of the main branch and only run the newly added packages. This will be harder if some packages are deleted.
On the other hand, I'm not sure this is worth the trouble. We always can rerun the checks once the package collection builds again. I'm planning to drop another note to the package maintainers of the currently failing packages over the weekend. If they require more time to fix this, the failing packages will be temporarily commented out.
I have no idea of how hard is it implement, but I've found the following:
When I propose to add some libraries to the collection, the whole collection is rechecked. So, when any library in the collection fails to build, the whole result of CI check for the PR is considered to be bad.
What I propose is to try to narrow the set of checked packages to changed/added packages and those packages that are anyhow influenced by those changes ones. This would lead to more adequate results of CI tests upon PRs.
The text was updated successfully, but these errors were encountered: