Professor Arms at Cornell once said during lecture that he was most impressed by the quality of the math articles on Wikipedia, and I think the same holds true for the computer science articles (and, really, when you get to the point of talking mostly about algorithms without code, that's what it really is anyway).
I find the maths articles on Wikipedia to be absolute and utter rubbish. This is because I can only understand them if I already understand the underlying material. But of course I am not going to be reading an article about something I already know, I'm going to want to read about something I (as yet) don't know.
The maths articles on Wikiepedia are -- in my experience -- useless at teaching people maths stuff.
Part of the problem with math in general is that it requires work to gain understanding. Describing something does not give you the intuition needed to work with the concepts.
Reading any article on math and expecting to understand the ideas is like reading a description of a language syntax and expecting to gain a feel for it. That simply doesn't work. Python, Lisp, OCaml, Haskell, they all require that you write something serious before you start to grok the language fully. The syntax is trivial, and the gains from understanding only that, likewise.
All articles on math need to be studied, not read. You need to work exercise, solve problems, use the material.
Otherwise all you have is a description, not understanding.
I try to get this point across to my dad (a physics professor) all the time. He is still hung up on its non-citability for scholarly purposes. But when you just need to look up something about the discrete wavelet transform and don't need to put a footnote anywhere, Wikipedia is just amazing.
This is fantastic... one central source for all that Wikipedia has to offer with relation to this subject... I always used to look up algorithms individually, without the knowledge of this central list. Thank you!
