So I've read through the code, and AFAICT this is all thats needed: - a unique ftesting.zcml per parallel runner, - with a unique +config-overrides.zcml file included - with a unique config name - which inherits from testrunner on disk. - and has a matching appserver config which inherits from testrunner-appserver
And then per-thing-that-needs-to-be-unique (e.g. dbname), we write that out into the unique config. Am I missing some stuff? Some context - we need to allocate unique ports, dbnames, etc for testing with, and while I'd like to do stuff really cleanly, I also want it working now :) - so I'm looking for the highest-leverage point to push on in the short term - whilst this might seem ugly, it seems like it should let us have the existing test code and config code manage to work. it would be nice if I knew why we had the -appserver variation though, I don't understand why its necessary or beneficial. -Rob _______________________________________________ Mailing list: https://launchpad.net/~launchpad-dev Post to : [email protected] Unsubscribe : https://launchpad.net/~launchpad-dev More help : https://help.launchpad.net/ListHelp
