You know, I head a lot of "Perl is dying" type stuff, and it really isn't. Version 5 is 10+ years old, and still being activly worked on. Bug fixes, performance enhancements etc. But the core language has stayed the same (STABLE!) for a long time.
It's just as useful today as it was. It isn't used much in the web world anymore because of CGI slowness (which has nothing to do with Perl), and mod_perl isn't available on a lot of web hosts.
I STILL contend that (comparing mod_perl to mod_php, or Perl via CGI vs PHP via CGI) that I can make the Perl version of something go just as fast, if not faster.
PHP is good for web sites because of what it is and how it was built. But it also has some serious shortcomings.
Perl 6... Been in development for what, 5+ years now? Not going to be out any time soon, and when it is it will be a serious difference in what Perl 5 was (though Perl 5 will still run just fine). It will address a lot of Perl 5's shortcomings. But will it catch on? No idea. For a lot of people, the learning curve may not be beneficial compared to what can be done in Perl 5, with not much effort. I doubt it will capture web space again, but you never know. If the whole Parrot VM thing takes off, who knows what direction things will move? Will be interesting.