Then there's the elusive release of Perl 6. Perl 5 was released in 1994, and in the past 16 years, the language has not had a stable full point release. Make no mistake, there's a vast amount of improvements in Perl 5.10 versus Perl 5.0, but there hasn't been much evolution beyond Perl 5 concepts. And it seems there won't be: Perl 6 isn't an upgrade from Perl 5 -- it's basically a different language. That new language hasn't had a stable compiler release yet, though the project has promised "a useable release of Rakudo Perl 6" on July 29.
This follows a decade of development, as Perl 6 was originally announced way back in July 2000. There has never been a road map or timeline for Perl 6's release, and given the stability and ubiquitous use of Perl 5, there wasn't a rush to release a new programming language based on Perl 5.
The new concepts, syntax additions, and myriad changes in Perl 6 may create a barrier to widespread adoption, especially with the vast amount of existing Perl 5 code out in the world right now. Where Perl 5 usage originally spread like wildfire, Perl 6 may not be nearly as well used, if for no other reason than Perl 5 is "good enough" and the intervening 16 years of Perl 5 code is baked into so many brains.
There are plenty of good reasons to use Perl 6, generally from a syntactic point of view. I'm a fan of the coherent sigil structures, where variable calls are constant no matter what the scope. Plus, object-oriented programming in Perl 6 is simpler and more robust (though much of that can be had in Perl 5 with Moose), and the concepts of roles and macros are definitely handy.
But I'm not sure these advances are enough to push Perl 6 into the limelight, let alone unseat Perl 5. There's also a much broader playing field now than in the early '90s when Perl really hit its stride. Today, Ruby, Python, and PHP are all extremely mature and capable languages.
I think of Perl as a Swiss Army Knife of IT, both Windows and Unix. In my opinion it's not a great Web language for a variety of reasons, including code complexity and performance, but if you need to, you can definitely write dynamic Web applications in Perl. It's not a great user interface language, but you can write a graphical UI in Perl. It's a fantastic text and string manipulation language, and it makes writing client/server applications extremely simple. It also has 20 years of modules that have been optimized over and over again, providing just about everything you could want in a programming language. Perl might not be the best tool for any given job, but it's almost certainly going to be able to do what you need it to do one way or the other.
That's Perl's strength: to be a jack-of-all-trades and a master of some. It's one of the pillars of the Internet explosion in the '90s and a foundation of modern computing. But has its time come and gone? I'd like to hear what you think.