I was a mathematics undergraduate in the MathWorld generation. It spread like wildfire in our department. I stopped carrying textbooks around with me – instead I could just walk into our undergraduate lab and look something up. MathWorld was every math textbook I needed. (A friend of mine was blocked from MathWorld after trying to download all the pages.) We mourned the year MathWorld disappeared.
Unlike my immediate reliance on MathWorld, I have been a slow Wikipedia adopter. The information on MathWorld seemed more reliable that Wikipedia could ever be, as it is contributed to (exclusively?) by mathematicians. That said, I can’t imagine Wikipedia going down for a year. (At least, not before the zombie apocalypse starts.) I find myself more and more using Wikipedia for technical matter. I don’t think it is solely because I am less likely to look up definitions of groups and more likely to look up definitions of complexity classes. I think the information on Wikipedia (even for MathWorld-type entries) is richer. Wikipedia entries tend to be more pedagogical than MathWorld’s, handy now that I am teaching.
I still don’t trust Wikipedia, though. It is a good, quick first reference; a source of examples. I’m hoping to incorporate Wikipedia participation into my graduate classes once I’ve figured out this teaching thing. Perhaps I’ll become more trusting of Wikipedia one day, but I wonder if I’ll ever rely on it.