On Fri, 2008-11-04 at 17:56 +0300, Shlomi Fish wrote: > > This is the first relevent one (subject changed and later in the new > > thread) > > > http://www.nntp.perl.org/group/perl.module-authors/2008/04/msg6404.html > > > > From what I can understand it reads data from META.yml and other > sources like > that, (or traps the output of Module::Build in a special mode.) While > such > ways to populate the machine-processable meta-data automatically may > be nice, > they are still not a direct substitute for having it in the first > place.
His claim, as I understand it, is that his code is will provide any meta-data you want to a search interface. It seems to me that what you want to do is to provide additional sources of such meta-data. What I want is to be able to find "best practices", standards, documentation, and appropriate modules from one place. I generally start with 'perldoc', then I try to guess and grep through the contents of .cpan/sources/*, then I try google and last I go to cpan. It seems to me that cpan should be higher up the list. > > That's because some aspects are better maintained in such a format for > the > convenience of the module writer and the programs (GUIs, WebUIs, CLIs, > etc.) > he runs to manipulate and verify them from the disk before they are > uploaded. -- --gh
