+1, bug is still there

Well, I've looked into this. If uri_decode is:

sub uri_decode { uri_unescape(shift) }

Then the url in the description works, but other scripts fail to display. If I change it to:

sub uri_decode {
    my $dec = decode('utf-8', uri_unescape(shift));
    Encode::_utf8_on($dec);
    return $dec;
}

Then other scripts work, but the above doesn't! I'm pretty confused, Unicode can be such a pain sometimes

Alright, seems like I was trying to be too clever here. Agreed solution is to go back to the classic behaviour - decode from UTF-8 (strictly), but if that fails don't try to do anything else (just display it as we have it in the database). For URLs that have pretty_names, they should not do anything special again, and just display the URL.

Are the changes made here related to MBS-3499?