openQA Project (public)

Observation¶

When there's a force_result softfailed with a carryover flag added, the first failing run does get the both the comment copied and the softfailed status is applied (see https://openqa.suse.de/tests/14672820#comments).

However, on the second run the comment is copied over but the status isn't applied, status stays as failed. This might be actually the wished for behavior, with the comment "The hook script will not be executed". One can then manually proceed to edit the comment in place without changes and clicking save, or if not the author of the original comment (cannot edit), copy pasting the same comment to another comment.

Expected behavior¶

Unsure, since carryover softfailed has dangers, but this should probably work logically one way or another. Maybe the needed fix woul dbe that the first time shouldn't auto-apply the status either either, because if RETRY would be 0 that could be unwished for.