Add running with falure status

[lunny]

No description provided.

[wxiaoguang]

I don't think it is well-designed and don't see why it is a must.

Please correct me if I was wrong: #34823 (comment)

[lunny]

I don't think it is well-designed and don't see why it is a must.

Please correct me if I was wrong: #34823 (comment)

There is a bug there. The job cannot be cancelled when it's running with failure because the aggregated status is failure not running. ref #35000 (comment)

[wxiaoguang]

Then it needs to detect the "cancel-able" basing on non-aggregated status?

[lunny]

Then it needs to detect the "cancel-able" basing on non-aggregated status?

However, it would still be strange if a failed task could be cancelled.

Alternatively, we could consider merging #35000. The waiting status, similar to running, should have higher priority than failure during aggregation.

[wxiaoguang]

Alternatively, we could consider merging #35000. The waiting status, similar to running, should have higher priority than failure during aggregation.

I think I have explained in #34823 (comment), job status and commit status indeed should have different state priorities.

correct me if I was wrong

[lunny]

Alternatively, we could consider merging #35000. The waiting status, similar to running, should have higher priority than failure during aggregation.

I think I have explained in #34823 (comment), job status and commit status indeed should have different state priorities.

correct me if I was wrong

#35000 is just for job status not for commit status.

[wxiaoguang]

Alternatively, we could consider merging #35000. The waiting status, similar to running, should have higher priority than failure during aggregation.

I think I have explained in #34823 (comment), job status and commit status indeed should have different state priorities.
correct me if I was wrong

#35000 is just for job status not for commit status.

Yes it is, so what's the problem?