in reply to Tried and True CPAN Modules

Here is one idea... Write some perl code that will examine CPAN to determine how many times a particular module is used by other CPAN modules and scripts.

Sort of a googlish approach to module usefulness or popularity, I suppose, but I think the results would be interesting.

It would be more effective to know how many times those modules are used outside of CPAN, but obviously that becomes problematic. Perhaps something that could search the archive of comp.lang.perl.misc to see how many times a particular module gets mentioned?

</ajdelore>

Replies are listed 'Best First'.
Re: Tried and True CPAN Modules
by Abigail-II (Bishop) on Jul 17, 2003 at 22:20 UTC
    That wouldn't give much useful statistics. It would at best give some measure of popularity, but not at all at usefulness. Some problems: there's no way to count how often the modules are used in programs, as programs are hardly uploaded to CPAN. You'd just count how often other modules use them. Furthermore, it would show a strong bias in favour of modules that don't have alternatives. For instance DBI. There isn't a module that does something similar, so the DBI would get high marks, even if the code itself is shitty.

    I wouldn't bother going to a site that ranks CPAN modules based on a popularity vote. After all, Windows isn't ten times better than Unix as an OS, is it? But that would be the outcome if you'd let people vote.

    I would value a site that does reviews of modules. Non-anonymous reviews, so a reviewer can establish a name for him/herself. Some monks are already exploring this idea after a similar discussion last week.

    Abigail

[reply]

In Section Meditations