I'm currently in the process of adding execution of nunit-console.exe to a NAnt script that will run regression tests after our build is complete. I need to distinguish three cases:
* All NUnit tests that were supposed to run did run and found no errors. I send out mail to the devleopers that all is well with the build.
* At least one test that was supposed to run did run and found one of the errors it was looking for. I send out mail to the developers that they bungled something (although more politely worded :-) ).
* "Something else" happened. Not all the tests that were supposed to run did get a chance to run or failed for some reason other than finding the errors they were looking for. I send out mail to the people responsible for the build process, for writing the regression tests, and for the NAnt script that runs the regression tests that they bungled it.
However, that isn't how the nunit-console.exe return codes are laid out.
* 0: All is well. That one I can use.
* 2: An internal NUnit error to be reported to the NUnit developers.
* 1: Everything else: A test found an error it was looking for, the tests never got a chance to run, the computer room is on fire, etc., etc.
If nunit-console.exe is going to return more than just 0/1 for pass/fail, then it should return exlicit (and DOCUMENTED) return codes of every condition it detects. In particular, each return statement in the nunit-console.exe code would have a distinct return code.
Merrill
------------------------------------------------------- This SF.Net email is sponsored by BEA Weblogic Workshop FREE Java Enterprise J2EE developer tools! Get your free copy of BEA WebLogic Workshop 8.1 today. http://ads.osdn.com/?ad_id=4721&alloc_id=10040&op=click _______________________________________________ Nant-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/nant-users
