On Thu, Jul 8, 2010 at 7:15 AM, Nigel Horne <n...@bandsman.co.uk> wrote: >> I can't help it if module authors do stupid things with their version >> numbers. > > That doesn't address the issue that I raised of informing developers that > they've made a mistake.
Historically, authors have complained about getting FAIL reports when prerequisites are not met so we have erred on the side of not sending FAIL reports in such a case. We *do* send PASS reports even if prerequisites aren't satisfied. I don't know of a general way to determine if authors mis-specified their prerequisites without an easy back-pan index to see if the version they specified ever existed and that would still presume an internet connection. In any case, that kind of prerequisite analysis is (mostly) static and I think it's better for CPANTS or some other static analysis tool instead of CPAN Testers. David