Stabilize CI for updated pull requests - #20
Conversation
Signed-off-by: Adam Jesionkiewicz <adam@jesion.pl>
ashtree74
left a comment
There was a problem hiding this comment.
Self-review completed. No blocking findings.
The check still enforces DCO on every non-merge contribution commit. It only excludes merge commits generated while updating a PR branch and uses the immutable PR author to identify Dependabot. Reproduction against PR #17 leaves its signed bot commit in DCO scope and excludes only the unsigned mechanical merge.
I cannot approve my own PR; after required checks pass I will use the documented maintainer PR-only bypass.
Signed-off-by: Adam Jesionkiewicz <adam@jesion.pl>
ashtree74
left a comment
There was a problem hiding this comment.
Follow-up self-review after the CI-discovered timing failure: no blocking findings.
The test now waits for the two observable synchronization events with bounded one-second expectations and one-shot callbacks. This removes both fixed RunLoop delays from the affected test without changing production code. The final affected test passed 10/10 after review edits (35 consecutive local passes total), and the full suite passed before the one-shot callback hardening. Both commits carry DCO sign-off.
Summary
Signed-off-byline on every non-merge contribution commitVerification
git diff --checkpassedSigned-off-by: Adam Jesionkiewicz adam@jesion.pl