Curating CPAN one dist at a time

curationCPAN Mon 20 January 2014

In a previous post I talked about deleting distributions from CPAN, and how there are plenty of dists that CPAN would be better without. In that I said I was going to start looking for candidate dists for removal. The first such dist, containing Win32::FUtils, has now been removed from CPAN (on 19th Jan 2014).

I started off by looking at distributions that were last released in 1996 (while working on the CPAN Report, I generated a list of dists sorted by the year they were last released). There were 22 dists, which you can see at the end of this post.

I worked through the list, looking up the dists on CPAN, checking to see if there were any dependent dists, their CPAN Testers stats, etc. I got to this release from Joe Casadonte:


I couldn't find the dist on MetaCPAN or; I looked at the author's directory on MetaCPAN, but nothing there. On, Joe's home page didn't list any dists, but it did mention the module that's in the dist: Win32::FUtils. In fact, the only way I could find the dist was by looking at Joe's CPAN directory:

When you look at the dist, you can see why it's not being indexed by the search engines: it doesn't follow any of the modern conventions for a dist.

I'm explaining all of this because to me that made it a good candidate for removal:

Joe's CPAN email address bounced, but after emailing a few other Joe Casadontes, I tracked down the right one, and emailed him. I explained what I was proposing and why. I got a positive message back from Joe, where he agreed that it could safely be removed from CPAN.

I emailed the PAUSE admins, copying Joe, asking for the dist to be removed. Ron Savage scheduled the deletion, which happened on Sunday night. Thanks to Joe and Ron.

CPAN dists last released in 1996

These dists saw their last (and probably first) release in 1996:

futils (J/JO/JOCASA/

I'll work through the others to see if there are any candidates for removal, or adoption, and then look at 1997.

I think other unreachable dists should be considered for removal from CPAN, and there are probably other types of dists that can be treated as candidates for removal. This should always be done respectfully: for the original author, any contributors, and other PAUSE users (who may be using them). The PAUSE admins take this kind of issue very seriously, so I see them as an extra guard / sanity check as well (while wanting to minimise what additional work I generate for them).

comments powered by Disqus