You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@maven.apache.org by Scott Babcock <sc...@hotmail.com> on 2022/05/11 01:27:32 UTC

Maven Surefire pull request #528

I’m requesting additional reviews and feedback on this Maven Surefire pull request<https://github.com/apache/maven-surefire/pull/528>, which resolves SUREFIRE-2075.

One of the reviewer expressed concern about the default value (5) which TestNG provides for the [threadCount] setting. I posted replies to this concern:


  *   I cited the fact that TestNG only utilizes this default value in parallel configurations in which no user-supplied value was specified.
  *   I pointed to the TestNG implementation that sets the thread count to 1 for sequential configurations.
  *   I pointed out that the current behavior of TestNGMapConfigurator (injecting a thread count of 1) conflicts with the standalone behavior of TestNG.
  *   I pointed out that this override results in sequential test execution for parallel configurations that don’t explicitly specify a thread count value.
  *   I made reference to the Maven Surefire documentation, which unequivocally states that the default value for [threadCount] is 5 if this setting is omitted.
  *   I pointed out that the existing implementation will throw a naked NumberFormatException for invalid [threadCount] values, whereas my revised implementation will return an actionable error.

If there are additional questions or concerns, I’m happy to respond.

Thanks!
= Scott Babcock =
https://github.com/sbabcoc

Sent from Mail<https://go.microsoft.com/fwlink/?LinkId=550986> for Windows

Maven Surefire pull request #528

Posted by Scott Babcock <sc...@hotmail.com>.
I’m requesting additional reviews and feedback on this Maven Surefire pull request<https://github.com/apache/maven-surefire/pull/528>, which resolves SUREFIRE-2075.

One of the reviewer expressed concern about the default value (5) which TestNG provides for the [threadCount] setting. I posted replies to this concern:


  *   I cited the fact that TestNG only utilizes this default value in parallel configurations in which no user-supplied value was specified.
  *   I pointed to the TestNG implementation that sets the thread count to 1 for sequential configurations.
  *   I pointed out that the current behavior of TestNGMapConfigurator (injecting a thread count of 1) conflicts with the standalone behavior of TestNG.
  *   I pointed out that this override results in sequential test execution for parallel configurations that don’t explicitly specify a thread count value.
  *   I made reference to the Maven Surefire documentation, which unequivocally states that the default value for [threadCount] is 5 if this setting is omitted.
  *   I pointed out that the existing implementation will throw a naked NumberFormatException for invalid [threadCount] values, whereas my revised implementation will return an actionable error.

If there are additional questions or concerns, I’m happy to respond.

Thanks!
= Scott Babcock =
https://github.com/sbabcoc

Sent from Mail<https://go.microsoft.com/fwlink/?LinkId=550986> for Windows