If Tim Gowers’s project for a wiki-like compendium of proof techniques catches on (and the search features work well), it could be a really big deal. These sorts of tricks are traditionally not recorded anywhere. Their use is buried deep in the folds of proofs where it’s difficult to see the scope of the idea. To learn the use of a particular technique you have to have the good luck to work with someone who knows the right tool and shows you it.
If used well, Tricki could speed up problem solving considerably. I don’t want to sound too excitable, but this is like the mathmos’ LHC.
btw. I lost my passport while I was in England and got stuck there (bad) and a bunch of other stuff is going on (good) so I doubt I’ll write the post I mentioned last time for a while. The moment has kind of passed.
EDIT: I could have sworn “It’s Tricky” went “It’s time to rock around”. Apparently it’s “It’s tricky to rock a rhyme” which makes more sense. I’m only half correcting the title though.