On Sun, Sep 19, 2010 at 09:19:21AM -0400, David Golden wrote: > On Sun, Sep 19, 2010 at 8:06 AM, Tim Bunce <[email protected]> wrote: > > Questions for cpantesters-discuss: > > - Are there any guidelines for minimum available memory for cpan-testing? > > Not that I've seen, nor would I support having them. The goal of CPAN > Testers is to test modules on as wide a range of platforms as > possible. That includes low-memory virtual hosts. Perl even > supposedly runs on android devices -- which would be a severe memory > constraint. I'm happy to have testers submit reports from as wide a > variety of systems as possible. > > If I get a new Kindle for the holidays, I might see if Jesse can help > me turn my old one into a dedicated CPAN Testers machine. :-)
Good point. It brings us back to the issue of failures due to the distribution being tested vs failures due to the platform it's being tested on. > > - Could the ENVIRONMENT AND OTHER CONTEXT section show available memory? > > The information reported is unique to each client (CPAN::Reporter and > CPANPLUS) though these have converged a bit in recent years. Ideally, > we'd have one common underlying module for providing such information. > > If someone were to offer up a cross platform way of reporting it, I'd > be happy to apply it to CPAN::Reporter. Looks like these two would cover the common Win32 and Linus systems: http://search.cpan.org/perldoc?Win32::SystemInfo http://search.cpan.org/perldoc?Linux::SysInfo Tim.
