|Perl: the Markov chain saw|
It sounds like you have been expecting to have magic happen, sstevens. Think about what you were trying to do. You want to have an OS set up, a webserver with an on-board embedded interpreter, AND a back-end interpreter behind that. Woof!
Just think about it for a bit. Even with Perl, setting up multi-layer web processing systems is no kiddie project. I agree that more of the rough edges have been sanded off with Perl, at least for mainstream deployments, but you still need to have some understanding to make it happen. Let's see: you want to set up an OS you don't understand, with a webserver you don't understand, with a module you don't understand, for a language you don't know, to do something you haven't figured out yet.
Sorry to be so rough on you -- even with > 27 years at it I STILL don't deeply understand all of the above -- but really, hubris can get you in trouble faster than anything. Larry Wall can have all the hubris in the world, because he WROTE Perl. Have you, by chance, studied the source code of any of the pieces you cavalierly tossed together? If you haven't, may I humbly suggest that you take one simple piece, like a symbolic LISP interpreter, and learn what it takes to make even that one little piece of the magic happen. There are dozens out there, even one written in Perl. Once you grok that fully, then maybe take a look at something a little more complex like the Ruby interpreter. Talk about awesome, there's some slick C code.
Your mistake is that you equate understandable with easy. Just because you can succeed in getting a Perl-centric webserver set up with less grief, does that imply that you understand it better? Does that imply that it IS better?
We all like to write nodes about goshwhata pearl is Perl -- I'm not immune either -- but let's go back and look at such nodes and use them to help us shore up our inadequacies. Deal? :D
"There's more than one level to any answer."