Keith:
> You might try a Unix-style way of doing it... use chroot(1) or
> chroot(2) to provide an environment within which the code can do what
> it likes, but be insulated from the rest of the system.
OK, I'll think about this.
Ideally I want the distinction (trusted / tainted code)
at the Haskell module (even identifier?) level,
because I want the test program to communicate with the test driver/logger,
which is a Haskell program as well, that *has* to do IO things.
With the above suggestion, they could only communicate
via the OS (by writing files).
regards,
--
-- Johannes Waldmann ---- http://www.informatik.uni-leipzig.de/~joe/ --
-- [EMAIL PROTECTED] -- phone/fax (+49) 341 9732 204/252 --
_______________________________________________
Hugs-Users mailing list
[EMAIL PROTECTED]
http://www.haskell.org/mailman/listinfo/hugs-users
