[PWGLF] Add QA histograms - #17440
Conversation
|
O2 linter results: ❌ 12 errors, |
alibuild
left a comment
There was a problem hiding this comment.
Auto-approving on behalf of @romainschotter.
|
Error while checking build/O2Physics/code-check for 21279dd at 2026-08-12 23:05: Full log here. |
alibuild
left a comment
There was a problem hiding this comment.
Auto-approving on behalf of @romainschotter.
alibuild
left a comment
There was a problem hiding this comment.
Auto-approving on behalf of @romainschotter.
|
Error while checking build/O2Physics/code-check for 35936a2 at 2026-08-17 12:12: Full log here. |
vkucera
left a comment
There was a problem hiding this comment.
The PR title is too vague. Please specify the scope of the changes (i.e. file or analysis). See the contribution guidelines.
alibuild
left a comment
There was a problem hiding this comment.
Auto-approving on behalf of @romainschotter.
|
Dear @vkucera, this PR is currently blocked because of your request for changes. However, the requested changes are based on an optional CI check. Could you please remove the blocking review and leave the corresponding comments as non-blocking feedback instead? |
Dear @mpuccio , if by "optional" you mean "can be ignored", then we don't have optional CI checks. The CI errors are input for the human reviewer to request fixing issues that a PR would introduce. Merging PRs with red CI sabotages this process, as was done in the initial PR that added this file. Please make sure that all LF code owners understand this. |
|
Dear @vkucera, no, by optional I do not mean that its findings should be ignored. They are certainly useful input for reviewers and code owners, and authors should be encouraged to address them. The point is whether findings from a non-mandatory CI check can, by themselves, be used to block a PR. My understanding is that this decision belongs to the relevant code owners, unless there is an agreed repository-wide policy making these checks mandatory. If this understanding is incorrect, I think we should clarify it with Analysis Coordination and apply the resulting policy consistently to all PRs. For this PR, I would therefore ask that the final decision on whether these findings are merge-blocking be left to the LF code owners (that would be us). |
Absolutely yes. That is the whole point of CI.
The policy is that code owners can decide to ignore some CI errors if there is a good justification for it. By default, errors should be fixed when they are spotted, if feasible.
In this case, the code breaks header dependencies and introduces uninitialised variables. These are absolutely merge-blocking. LF code owners should not have the impression they can ignore these kind of errors. This is why I blocked the merging. |
|
Dear @romainschotter , @mpuccio and @vkucera , |
|
Dear @dsekihat ! |
|
Error while checking build/O2Physics/o2 for 35936a2 at 2026-08-17 15:14: Full log here. |
No description provided.