Beefy Boxes and Bandwidth Generously Provided by pair Networks
Perl Monk, Perl Meditation

CPAN Reference Counter II

by pileofrogs (Priest)
on Feb 04, 2010 at 01:08 UTC ( #821295=perlquestion: print w/replies, xml ) Need Help??

pileofrogs has asked for the wisdom of the Perl Monks concerning the following question:

This post is an attempt to clarify and focus my earlier post: CPAN reference counter?.

I would like to know how many time a module gets used by other modules. There seems to be a few projects which attempt to figure out what modules a module depends upon, which is almost but not exactly the opposite.

The reason I want to know this is to have a kind of popularity/usefulness metric. I'm interested in helping people who are trying to decide what module among many to use when confronted by an array of choices. It would also provide an easy top-ten list for newbies looking to familiarize themselves with perl.

In academia, a lot of scientists count the success of a given article by the number of other articles that reference that article, and I think the same thing could apply here.

Now, here's the real question (and why I'm posting again...): Does this have to be exact? I don't think it does. If it's used by humans trying to whittle down a list of possible solutions, it can be a little out of date and miss a few references without ruining anyones day. In that light, it's like kwalitee: useful but not the end of your problems.

Am I crazy? Does this sound like a good idea?



Replies are listed 'Best First'.
Re: CPAN Reference Counter II
by toolic (Bishop) on Feb 04, 2010 at 01:53 UTC
    Does this have to be exact?
    Am I crazy?
    Maybe :)
    Does this sound like a good idea?

    I think it would be interesting to see the results, and even more interesting to see the Monks' opinions of the results.

    Here are some expected outcomes:

    • Core modules and pragmas would tend to be used more often than non-core.
    • Older modules would tend to be used more than newer ones.
    • Smaller modules would tend to be used more than larger ones.
    • General-purpose modules would tend to be used more than modules that target a specific use.

    If a module is used by many other modules, then there is a good chance that the module is of high quality. I would expect Carp to be used by thousands of modules, and it is a good module, in my opinion.

    However, I would expect that another great module, like XML::Twig, would not be used by many modules. I am not trying to compare this to Carp, because they are in totally different categories. My point is that some great modules might not be used by any other modules. Just because XML::Parser is used by 100 modules, and XML::Twig is used by 2 (or 0), doesn't mean a user should choose XML::Parser over XML::Twig.

      Precisely! If you're trying to figure out which XML module you want to use, you could compare the scores of the different XML modules you are considering.

Re: CPAN Reference Counter II
by JavaFan (Canon) on Feb 04, 2010 at 12:02 UTC doesn't answer your question?

      Not really. First, it does distributions, not modules and second, it doesn't do core. You can't get a used-by list for 'CGI' or 'Carp', for instance. Also, it's ungainly for my purposes. You'd have to enter each distribution you're considering individually, then count the number of used-by distributions to get a score. Lastly, it doesn't provide a 'top 10'.

        Doing distributions instead of module, I'd consider it a feature. Or should a module that's used in a single distribution in 100 different modules count 4 times as heavily as a module that's used once in 25 different distributions?

Log In?

What's my password?
Create A New User
Node Status?
node history
Node Type: perlquestion [id://821295]
Approved by toolic
and the web crawler heard nothing...

How do I use this? | Other CB clients
Other Users?
Others perusing the Monastery: (5)
As of 2021-02-26 15:57 GMT
Find Nodes?
    Voting Booth?

    No recent polls found