Here's the thing: to some extent, you're right, but at the same time, you're still way off base. Now let's say, just for a second, that the only thing Perl can / should be used for is system administration and web programming (please don't flame me, guys. I know that is by no means the extent of Perl's capabilities, but I don't have time to write a post covering all the bases right now.) Even assuming that's true, you're still wrong.

   Considering the web side of it, yes, Perl is being used in far fewer places now. There are two major reasons for this: one is the abundance of other languages (PHP and ::shudder:: ASP, for example). Another is the fact that a lot of sites backed by Perl programming crashed and burned when the dot-coms did. You know what? I don't see this as a big deal. The techies who wrote those sites are still around, likely still using Perl, and hopefully not hurting its reputation by using it to support companies that will never, ever, make any money or do anything useful...ever. (Not to say that all dot-coms were this way, but c'mon, there were quite a few useless companies out there.) These sites were thrown together (in many cases) to make a quick buck. Granted, Perl can be great for quick-and-dirty code, but do we really want it making up the majority of the volume of the code out there?

   System administration: I still think Perl is one of the finest languages ever created for system administration, especially cross-platform system admin, for those of us who don't want to learn the ins and outs of shell scripting 100x over. I really don't think there'll be much argument there, so I'll move on.

   The community: when was the last time Perl had organizations as devoted as the Perl Mongers and Yet Another Society? Do you see a website as popular and helpful as Perlmonks for PHP? Before last year, I don't remember ever having programmers whose sole job is to improve and evangelize Perl. Do you? I can't remember ever before having an argument on the Linux kernel mailing list on whether Linux should switch to a quasi-Perl method of patch management. (This past week, for those who haven't been reading the kernel mailing list or Slashdot.)

   To be honest, I have no idea what you're talking about. As far as I'm concerned, Perl is as strong as it's ever been. If not, then it's purely a matter of evangelization. I know I've impressed more than one boss by getting what would've been a several-day-long job done in an hour with Perl. Have you?