I understand what you are saying. You don't want to rely on prepackaged solutions so much that you don't have the opportunity to learn and to improve. That's a good attitude to have.

On the other hand, if you roll your own recursive file finding algorithm or (and you all knew this was coming) a CGI handler, you'll have something that works in 80% of the situations with very little effort. Unfortunately, the other 20% of the time, it'll take ten seconds for someone in the know to point out where it breaks. I don't want to spend the next week reading three or four RFCs and guessing at what's broken in a sea of web browsers, robots, servers, standards, angels and daemons.

At some point, both of us have to get our work done. That doesn't mean that I didn't take apart File::Find the other night, trying to put a nicer interface on it, or that jlp didn't rip out most of CGI so we could make it sing and dance a little better. Heck, some people here have even read the Perl source code in spots. (Other spots I wouldn't recommend.)

I'm all for learning. Download the module. Read the code. Figure out what it's doing and why. I'm just not going to advise people that before they get any work done they need to build their own computer, install Linux From Scratch, delete Configure.sh and write their own Makefile before they can feel comfortable downloading something like LWP.

There may be a handful of third graders who could write a passable sestina, but let's give them a chance to copy stuff from the chalkboard before we push them too hard.