the regex implementation is extremely slow, but in general it's quite reasonable. i wouldn't use it for number crunching, but for all the things that we've been

the regex implementation is extremely slow, but in general it's quite reasonable. i wouldn't use it for number crunching, but for all the things that we've been using perl/ruby for it seems much faster (and it has jvm concurrency, which means we can thread really nicely which helps too). i'm going to be at web 2.0 next week but when i get back i'll put together a tarball that has some examples and stuff in it and send it along. the project is called 'jetstream' by the way. please send me an email to remind me. cheers, ben