James Michael Dupont <[EMAIL PROTECTED]> wrote: > Dear fellow hackers,
> First of all, I would like to say that I hope that you dont take > offence at my silly questions and crazy ideas! ;) There are no silly questions, only silly answers and crazy ideas are the motor of development - trying to answer some: > 1. Is it possible to stored in some way ASTs inside of parrot IMC byte > code in such a way that they may be evaluated and compiled at runtime? Not yet. But parrot should finally be able to run from some kind of AST directly. > 2. Is it possible to emit new IMC routines on the fly and have them be > loaded in as if they came from the load routine? like a form or > reflection.emit? Yes. Have a look at t/syn/eval.t and languages/parrot_compiler/parrot.pasm. > 4. Is there any attempts at debugging support in parrot, for a way to > mix the original source code and assembly together, a way to enable > tracebility of. HLL source is currently being discussed. PASM source + i386 assembly[1] can be stepped through: s. docs/jit.pod and docs/debug.pod. [1] may be other processors too. > mike leo
