Tibor17 commented on PR #479:
URL: https://github.com/apache/maven-surefire/pull/479#issuecomment-1152928266

   When I see the 
[commit](https://github.com/apache/maven-surefire/pull/479/commits/a8295bfebd4c067d4caa1323ae80ecfee42883ac),
 I have to say that this PR is a pure mess and disaster, the providers must not 
be touched in the ITs POM. The POM changes have nothing to do with any build 
speed up, it has nothing to do with `-T3`.     It's a pure hack which affects 
functionality of the tests and it is very risky solution.
   As I said before in Slack or GH, the way how the build can be faster is to 
focus on the most slow IT classes and those are only few and those are 
Parameterized and they can be speed up as it was before (I deleted due to Maven 
Resolver) only the way that the Maven local repo would have all artifacts  
pre-loaded or the Resolver would be failsafe which is not now and so the 
concurrent ITs would fail the Resolver of course. Maiking the project build 
parallel would speed up the build in 5 min because all modules except the 
surefire-its are much faster in comparison to the ITs, so there is no reason to 
think about these "improvements". The penalty would wrong behavior of the 
tests, and that's worst!
   
   Another story is embedded Maven. This was not stable solution for us some 
time ago because it shares system props and env is shared and the developer has 
no notion about while writing the test, and so there is the risc the IT however 
succeeds on Jenkins but fails in real environment. That's the reason why it was 
removed several years ago in the surefire-its/pom.xml.


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to