Alceu Rodrigues de Freitas Junior <[email protected]> writes: > Greetings, > > I have some doubts about using distroprefs and I could not find > information about it: > > 1 - Once I setup a distroprefs to skip a distribution (because it > hangs during execution, for example) the maintainer of the > distribution actually receives some information telling the > distribution was skipped to be tested? I believe that the maintainer > should received some feedback (even if the only possible thing is to > use something like Devel::AssertOS to skip unsupported OS);
He does not get feedback. And I think one of the reasons to skip distros is not the fail but that some maintainers explicitely do not want to get any feedback. > 2 - There is any way to keep control about when to stop skipping a > distribution? I know that regular expressions could be used to not > test a specific version, but when using a smoker probably some > processing hours will be lost when you decide to take a look and see > that the same distribution is hanging there forever. Distroprefs are a static thing. Orthogonal to a smoker. There is no auto-finetune. > 3 - Is there any plan to create a single repository for distroprefs? > We are probably losing some synergy by using our own or grabbing some > from other testers. I maintain a central one with git submodules to all other repos I'm aware of: https://github.com/renormalist/cpanpm-distroprefs Though, I didn't update the submodules for a while... Tell me yours and I add it, or send pull requests, the usual drill. For my purpose I then merge them as needed, usually use "cpanpm" as base and overwrite some with my own "renormalist" tweaks. Kind regards, Steffen -- Steffen Schwigon <[email protected]>
