Ever wonder which CPAN mirror in your country is the most up to date? Just run this script, passing your two-letter country code on the command line, defaulting to 'us'. Requires XML::XSH2. The resulting output is sorted by age in hours. For Germany right now, we get:

ftp://ftp.uni-erlangen.de/pub/source/CPAN/ => 1
http://www.planet-elektronik.de/CPAN/ => 1
http://cpan.noris.de/ => 2
http://netmirror.org/mirror/CPAN/ => 2
ftp://cpan.mirror.iphh.net/pub/CPAN/ => 2
http://www.chemmedia.de/mirrors/CPAN/ => 3
ftp://ftp.fu-berlin.de/unix/languages/perl/ => 3
ftp://ftp.rub.de/pub/CPAN/ => 3
http://ftp.uni-erlangen.de/pub/CPAN/ => 4
ftp://netmirror.org/CPAN/ => 5
http://cpan.mirror.iphh.net/ => 6
http://cpan.provocation.net/ => 8
ftp://cpan.provocation.net/ => 8
ftp://cpan.noris.de/pub/CPAN/ => 10
ftp://ftp.gwdg.de/pub/languages/perl/CPAN/ => 11
http://dl.ambiweb.de/mirrors/ftp.cpan.org/ => 11
http://ftp.gwdg.de/pub/languages/perl/CPAN/ => 11
ftp://ftp.mpi-sb.mpg.de/pub/perl/CPAN/ => 11
http://www.couol.com/cpan/ => 12
ftp://ftp-stud.fht-esslingen.de/pub/Mirrors/CPAN/ => 12
ftp://ftp.cw.net/pub/CPAN/ => 13
http://ftp.cw.net/pub/CPAN/ => 13
ftp://ftp.gmd.de/mirrors/CPAN/ => 13
ftp://ftp.cs.tu-berlin.de/pub/lang/perl/CPAN/ => 13
http://cpan.triplemind.com/ => 14
ftp://ftp.freenet.de/pub/ftp.cpan.org/pub/CPAN/ => 15
http://cpan.mirroring.de/ => 15
ftp://pandemonium.tiscali.de/pub/CPAN/ => 16
ftp://nlp.petamem.com/ => 16
http://pandemonium.tiscali.de/pub/CPAN/ => 16
http://cpan.interprovide.de/ => 21
http://www.mirrorspace.org/cpan/ => 28
http://www.cpan.mirroarrr.de/ => 672
http://cpan.linux-mirror.org/ => 2224.8
[download]

Yes, that means that last mirror hasn't been updated in three months!

This code is for an upcoming Linux Magazine column.

#!/usr/bin/env xsh
quiet;

# open --format html 'cpan.html';
open --format html 'http://www.cs.uu.nl/people/henkp/mirmon/cpan.html'
+;

perl { $country = @ARGV ? shift(@ARGV) : 'us' };

for //tr[th/a/@name = $country]/following-sibling::* {
  if th/a/@name last;
  $url = (td[1]/a[1]/@href);
  $time = (td[3]);
  perl { $time{$url} = $time };
}

perl {
  for (values %time) {
    s{([\d.]+) days?}{$1 * 24}e
      or s{([\d.]+) hours?}{$1}
        or s{([\d.]+) minutes?}{sprintf "%.1f", $1 / 60}e
          or die "Don't understand $_";
  }

  for (sort { $time{$a} <=> $time{$b} } keys %time) {
    print "$_ => $time{$_}\n";
  }
}
[download]