Re: compiler directives; (was: Re: cheerful static typing (was: Any and Every... (was: Eval)))

[Noel Welsh <address@hidden>]

--- Michael Vanier <mvanier@cs.caltech.edu> wrote:
> 
> This brings up an interesting point.  How do people
> on this list feel about
> allowing compiler directives in programming
> languages?

I'm in favour of compiler directives. I believe there
should be two-way communication between the programmer
and compiler.  I should be able to say "this imported
C function is a pure function, so hoist it out of
loops".  The compiler should be able to say "I can't
do some neat unboxing here because I can't prove the
types of this expression.  Can you simplify it?"  And
so on.  The Python (Common Lisp) compiler does the
latter but few systems do the former.

A search on google for compiler annotation brings up a
lot of hits, so there appears to be a lot of work in
this area.  Two papers I'm familiar with describe
annotations in C-- and MLRISC.  My LtU post was
vaguely about this topic (but fairly vague about any
topic):
http://lambda.weblogs.com/discuss/msgReader$2161

Noel


__________________________________________________
Do You Yahoo!?
Send FREE video emails in Yahoo! Mail!
http://promo.yahoo.com/videomail/