[
https://issues.apache.org/jira/browse/FLINK-1166?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14307600#comment-14307600
]
ASF GitHub Bot commented on FLINK-1166:
---------------------------------------
Github user mxm commented on the pull request:
https://github.com/apache/flink/pull/366#issuecomment-73087911
I just ran the script on your pull request ;)
> Flink QA-Check results:
> :+1: The number of javadoc errors was 402 and is now 402
> :-1: The change increases the number of compiler warnings from 104
to 314
> :+1: The number of files in the lib/ folder was 142 before the
change and is now 142
> Test finished.
> Overall result: :-1:. Some tests failed. Please check messages above
I guess if the user does not rebase to the latest master, then negative QA
test results are likely to occur. Also, warnings could be hidden when the
number of warnings have increased on the master. Wouldn't it be better to
compare against the base of the user's changes?
> Add a QA bot to Flink that is testing pull requests
> ---------------------------------------------------
>
> Key: FLINK-1166
> URL: https://issues.apache.org/jira/browse/FLINK-1166
> Project: Flink
> Issue Type: New Feature
> Components: Build System
> Reporter: Robert Metzger
> Assignee: Robert Metzger
> Priority: Minor
> Labels: starter
>
> We should have a QA bot (similar to Hadoop) that is checking incoming pull
> requests for a few things:
> - Changes to user-facing APIs
> - More compiler warnings than before
> - more Javadoc warnings than before
> - change of the number of files in the lib/ directory.
> - unused dependencies
> - {{@author}} tag.
> - guava (and other shaded jars) in the lib/ directory.
> It should be somehow extensible to add new tests.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
