I have a few ideas regarding tests on Windows and maybe other OS-es: 1) When looking at the matrix (e.g. http://matrix.cpantesters.org/?author=szabgab ) the FAILs stand out as they are red but the "no tests on this platform" is a less obvious thing. The white column stands out in many cases but it is a neutral thing instead of a warning sign. I am not sure if a warning sign should be there or not and if Windows should be special case compared to the other less frequently used (and tested) platforms.
2) Maybe we should collect the "disabled" files and display their content in a central place. That might help seeing the difference between "This module was not lucky and got no or few reports" and between "Testers don't test this module for some reason". If this could be integrated with dependencies as well that would be even better so we can see. That way an author could look up his modules and see if any of them or any of their dependencies is on the black lists? 3) A report on the relative number of reports of each distribution. The problem is that when I see a distribution that has 30 reports I don't know if that's a lot or a few. So I don't know if, as above, it was just not lucky or if there was some other reason it did not get test reports. Certainly the time the module was uploaded to PAUSE can have a huge impact on the test reports. A 1 day old package can expect to have a lot less reports than one which has been on CPAN for 2 weeks so this needs to be taken in account somehow. Anyway, these are just ides for discussion. Gabor
