npm, PyPi overtake CPAN

[Nicholas Clark]

On Fri, May 24, 2013 at 11:59:03AM +0200, Mallory van Achterberg wrote:
> I was at the PyGrunn meeting in Groningen recently.  Holger Krekel 
> (https://twitter.com/hpk42) gave a talk about the problems of pypi
> and compared it to CPAN, and then showed some new work he was doing
> to help fix the multiple issues. He loves CPAN's searching, testing,
> multiple mirrors etc.

Mirrors matter. (Well, mirrors are important, but that doesn't have the
alliteration) because

1) in an online environment, just a URL to a master site is a single point of
   failure. Start to chain enough dependencies, and however reliable they are
   at building and testing, it doesn't work if you can't download the code
   because one server is down
2) in a secured environment it is easi*er* to make a mirror and ship it in
3) Sometimes I am on a $DHH plane. Or in a tunnel. Or in
   data-roaming-multiple-limb-loss land. So having a mirror is useful.

> He has slides but I haven't seen them published anywhere, but the 
> praise for CPAN was great and mighty, and it's good for Python to
> see someone basically wants to make the Cheeseshop more like CPAN :)

Yes, it nice to know that Perl is doing at least some things right.

I was also surprised to notice that the new (<8 weeks old) official mirror
client for PiPy, bandersnatch, is written in Python 2.7. "Official", because
the description and link are from here https://pypi.python.org/mirrors to
here https://pypi.python.org/pypi/bandersnatch

Nicholas Clark