Volunteer! Time series stats of new PAUSE ids

I've finally realized that instead of people worrying about false, or ever true, reports of Perl's demise, we can generate some factual (nice change!) data about Perl activity.

So, I'd like to see someone analyse the postings to the archive of modules at perl.org. See e.g.:


The only thing you'd need to look for would be Subjects of the form:

Welcome new user XXX

A time series of say # of new ids per month for the last 10 or even 20 years would give us something to talk about.

Would it be proof of anything? No :-(, probably not :-).


I have most of the data in pause-replay and you could steal my job counter script from jobs counter.

A lot of people sign up but never upload a module. More interesting is the number of first-time uploads per month. By a strange coincidence I generated that graph yesterday, for a blog post on an idea.

I haven't finished playing yet, but here is a graph of the number of PAUSE ids uploading at least one thing, by month.

The numbers for Q4 of 2013 look worryingly low. ;-)

New modules uploads vs Updates for existing modules would be interesting.

There was an article somewhere (cannot find a link), with some figures showing that updates/new relation for CPAN is somewhat equal to new/updates for Rubygems.

Victor: I'll knock something up in the next day or so, unless someone beats me to it.


found a link: http://blogs.perl.org/users/hashbangperl/2013/03/comparing-apples-and-oranges---rubygems-vs-cpan-part-1.html

When you look at the numbers of recent uploads, 42 out of the most recent 50 (which is what the rubygems api gets you) are first releases (possibly more, I gave them the benefit of the doubt), while cpan uploads are almost the reverrse at 9 out of 52.

Ah you mean this graph that's been a part of CPAN Testers Statistics from quite a few years ;)


Click on the graph to see the trends over the years.

Leave a comment

About Ron Savage

user-pic I try to write all code in Perl, but find I end up writing in bash, CSS, HTML, JS, and SQL, and doing database design, just to get anything done...