In the following example I keep getting "1: Forbidden" as output, but I don't see what is the real difference between the two calls. Can anyone see it? UIR::Fetch is at 0.09 and LWP::UserAgent is 6.04.

use strict;
use warnings;

use URI::Fetch;
use LWP::UserAgent;

my $source = 'http://thorsen.pm/atom';

my $res2 = URI::Fetch->fetch($source);
print '2: ' . URI::Fetch->errstr if not $res2;

my $ua  = LWP::UserAgent->new;
$ua->env_proxy; # force allowing of proxies
my $res1 = URI::Fetch->fetch($source, UserAgent => $ua);
print '1: ' . URI::Fetch->errstr if not $res1;
[download]

If I change the $source to be http://code.foo.no/feed then I keep getting "1: Bad Behavior".

Very strange. Other URLs I tried work well.

And just to explain why is this important, I am using XML::Feed to fetch and parse RSS and Atom feeds, and XML::Feed uses the code as it is written in the problematic case. So currently these two feeds keep giving me error.

Update

I found it. I think. Earlier, when I stepped through the code with the debugger I missed it, but now I saw the difference. The LWP::UserAgent created by URI::Fetch identifies itself as "URI::Fetch/0.09" while the default UserAgent I passed to the fetch method had the default identification string 'libwww-perl/6.04'.

Apparently those two servers refuse to serve to the well known Perl user agent. Probably in an attempt to fend off certain bots.

So I have to change the UserAgent in XML::Feed.