I have had a theory for a long time that perl based CGI scripts could be sped up considerably by simply by keeping the interpreter resident in memory. Not only that, but the module that makes that possible wouldn't be a pain in the ass to install because of a key design goal: simplicity, borrowing heavily from a module that I think does a lot right, mod_php.
Well, if you followed the thread you would have noted that Aaron just recently got a version of mod_perlite to work against basic CGI functions. Totally awesome.
Next step: benchmarking. There are some skeptics out there, and those skeptics are a helluva lot smarter then me, so I have to see for myself: does this approach have actual merit?
And if the theory holds water, then I will setup a homepage for the project, get some builds spun up, and start working on documentation. For those of you who expressed interest in helping, by all means check out the latest from subversion and give mod_perlite a whirl.
This is an exciting project. I have hoped for something that blended some of the performance benefits of mod_perl with the ease-of-use and lower-persistence memory footprint of PHP. It sounds like mod_perlite could be this kind of solution.
That is very cool. If it works out maybe hosting companies will start using it by default like they do PHP and I will be a very happy Perler.