[Enhancement] Removing the Bazel workflow
Before Creating the Enhancement Request
- [x] I have confirmed that this should be classified as an enhancement rather than a bug/feature.
Summary
Previously, we introduced the Bazel workflow in our community, which can quickly provide test results, but it also introduced some issues:
-
The maintenance cost is relatively high. This includes timely updates to configuration files when new packages are introduced. The workflow often fails due to configuration issues, which can be confusing even when there are no actual problems.
-
Although it can quickly provide test results, multiple workflows run simultaneously, and due to the bucket effect (wooden bucket principle), the final test results still require a long waiting time. Moreover, the test cases in other workflows can cover those in the Bazel workflow.
Therefore, I would like to propose removing this workflow. If you have any suggestions, please feel free to reply below.
Motivation
Removing the Bazel workflow
Describe the Solution You'd Like
Removing the Bazel workflow
Describe Alternatives You've Considered
No
Additional Context
No response
Can be deleted, this workflow is often unstable.
Please assign it to me,I would like to work on this.
Please assign it to me,I would like to work on this.
done
Can be deleted, this workflow is often unstable.
You guys get things totally wrong. It is not the workflow not stable, it's the unit tests that are flaky.
“ The maintenance cost is relatively high. This includes timely updates to configuration files when new packages are introduced. The workflow often fails due to configuration issues, which can be confusing even when there are no actual problems.”
Given the frequency of adding new dependency, this reason does not hold.
The workflow offers resource tight scenarios, exposing unstable, flaky test cases and potential bugs... often not found in the hours long slow workflow.
Can be deleted, this workflow is often unstable.
You guys get things totally wrong. It is not the workflow not stable, it's the unit tests that are flaky.
I had a PR to solve this problem. I found that there was a difference between jdk8 and jdk11. jdk8 can pass, but jdk11 cannot. In addition, I noticed that there was an issue to support jdk11. Is this still going on? I want to participate?