Hello,
I'm not that familiar with python, but in any case ... This is the
academic project involving 'extensible systems', where routines are
allocated in some memory shared among the multiple address spaces, and
user process can call those routine via kernel system call. One of the
key requirements for those routines is 'to be safe'. At this point I'm
looking as safe extensions to C (such as Cyclone), but at this point I
ask myself: why not an interpreted language like Python? I guess it is
safer than C and yet, as I undestand, is fairly powerful language.
What would be required is to have the whole thing in a core memory.
Either somehow compile python script, so it becomes the loadable
module (something similar to GJC, Gnu Java Compiler), or may be having
Python interpreter sitting in a core, and Python script compiled as a
bytecode, and also somehow loaded int the core. Would like to hear
from those familiar with Python internal workings if this is idea
worth persuing.
TIA.
