Repository navigation
Conversation
|
O2 linter results: ❌ 0 errors, |
|
Hi @Rrantu , please fix the errors and warnings. |
|
Error while checking build/O2Physics/code-check for 56de32c at 2026-10-19 10:42: Full log here. |
|
Error while checking build/O2Physics/code-check for d97bf67 at 2026-10-19 13:39: Full log here. |
|
There is still a warning from O2 linter. |
Hi @vkucera , it seems that I can’t resolve these two warnings within this PR, as both would require changes to the upstream workflow or repository settings. |
|
Error while checking build/O2Physics/code-check for 821e08a at 2026-10-19 17:03: Full log here. |
|
Error while checking build/O2Physics/code-check for 4b5a28f at 2026-10-20 18:58: Full log here. |
Hi @Rrantu , thanks for the effort but I was referring to the O2 linter warning. I assume you tried to fix the code check warnings instead. |
Hi @vkucera , the O2 linter warning seems to be related to .github/workflows/o2-linter.yml and Node.js 20. I’m not sure if there’s anything I can do on my side to address it. |
That is not an O2 linter warning. The linter log is always in the "Run tests" step. The linter cannot analyse the |
|
|
||
| #include <cstdint> | ||
| #include <vector> | ||
| #include <vector> // IWYU pragma: keep |
There was a problem hiding this comment.
Please do not add the pragma. This is a temporary issue that I should be able to fix soon.
Thanks a lot! I've fixed this. |
|
Error while checking build/O2Physics/code-check for 8814094 at 2026-10-20 18:57: Full log here. |
Extend the Xic0 to Xi π KF ML response to support the 28 input features used by the local PbPb Transformer model.