jvm bytecode was designed to be fairly fast to int...

# announcements

[mikehearn]

jvm bytecode was designed to be fairly fast to interpret and JITC, LLVM bitcode was not. they make different tradeoffs. the latter is a good call for building a set of C/C++/ObjectiveC compilers. the former is a good call if you want to be able to upload a JAR and have it be run on all kinds of machines including types that haven't been invented yet, with speculative optimisations