I am withdrawing from Wikipedia editing indefinitely. It's not worth the time or the stress. I may return at some point — who knows.
You know what? I give up. The return on investment of trying to make Wikipedia an encyclopedia instead of a multi-subject fansite just isn't worth it. So I'm going away to get real work done someplace where I won't have to spend more time fighting cruft than adding content. For those of you who do still care, here are my thoughts on how it could be improved.
I've noticed that Category:Computer science articles tend to have a strong bias towards object-oriented programming, imperative programming, and lazy evaluation. Wikipedians tend to write about specifics of Java, C++, C#, Python, Perl, Scheme, and Haskell, often to the exclusion of the more generic discussion one expects in an encyclopedia. Real-world examples are good, but over-specifying is bad; articles should make sense to people who don't already know a particular programming language. See WP:NPOV and WP:AXS.
Speaking of over-specifying, lists on Wikipedia are usually not a good idea; there's a reason why actual print-encyclopedias don't contain very many of them. A "list of things having this attribute" is almost never complete, and makes maintenance very difficult. Because a list usually misses a lot of important items, it almost always carries the POV of whoever added the items that are there; marking all lists as stubs would help there, but isn't really practical. Categories, on the other hand, are designed to be dynamic, and WP:CATP is helping make them better. See Wikipedia:Lists#Appropriate topics for lists, WP:CG, and WP:CLS for more details.
Candidates for deletion
I'm not a deletionist, but I do support keeping Wikipedia an encyclopedia (see WP:NOT). This is why I'm leaving — there are far, far too many articles that are zealously defended by people who believe that because Wikipedia is not paper it should have content ranging from product guides to running debates to lists of unverified programs that all do the same thing. Non-paper doesn't mean non-encyclopedia, after all.