OpenACS is a bit of a beast. AOLserver is a very nice bit of work, though. It's a pity it wasn't released as open source sooner, as it is a very nice system for creating dynamic content.
Agreed. I really enjoyed AOLServer. It was rock solid, really fast and reasonably resource efficient, did everything we needed for our particular deployment, and was reasonably well-documented.
But, it also had some limitations that made it less promising. It was intimately tied to Tcl, which is, let's be honest, not growing in popularity, had some old school architectural choices that would limit extensions and be hard to overcome (Apache had a lot of the same issues early in its life, but overcame them through brute force...when there are hundreds of active developers, all problems get smaller). It was designed for a very specific purpose, and it showed. So, I don't think it's a codebase I would consider spending time or effort on, if I didn't have really strong reasons for sticking with it.
