A good place to start is with the library loading code. This is in the works, and something we need, but we also need to figure out *how* it should work. Should it, for example, override the built-in load_bytecode op? Or should it instead be presented as a regular parrot sub, in which case what should it be called and what are its semantics?
We also need to decide how the library itself will be loaded up. (Given that the library has the code to find and load libraries there's something of an interesting bit of fun possible there) My personal preference is to have the bytecode for the library in an OS-level shared library that parrot links against for easy upgrading, but that makes building parrot in the first place somewhat problematic. (Which would be solved some if we had miniparrot working... nudge, nudge! :)
Anyway, now's the time to weigh in on this, since once we nail these bits down I think we'll spin off a separate parrot-stdlib list to deal with the all-bytecode standard library. (And I'm still trying to get a parrot-compilers list for compiler writers. Some day, I hope...)
--
Dan
--------------------------------------"it's like this"-------------------
Dan Sugalski even samurai
[EMAIL PROTECTED] have teddy bears and even
teddy bears get drunk
