On Sat, Sep 5, 2009 at 3:28 PM, John Cowan<[email protected]> wrote:
> Ray Dillinger scripsit:
>
>> This would involve rules about what (shared) variables are "in scope"
>> in particular threads, with other threads not permitted to use them,
>> right?  Or with particular threads having "write" access and other
>> threads having "read" access to the same variable?
>
> Unfortunately, shared variables aren't the half of it.  Shared mutable
> data structure is a much worse problem, because the compiler can't
> catch it.  If I pass the same list (in the sense of eq?) to two threads,
> and each of them calls reverse! on it with the classical constant-space
> algorithm, my program will become "free and wild and beyond good and
> evil, with laws and morals thrown aside" (HPL, "The Call of Cthulhu"),
> and (more to the point) unlikely to do the right thing.
>

Not to diminish the difficulty of the problem, but I believe you can
reverse the direction of closure conversion to reduce the general
data structure problem to one only involving mutable variables.

That is, I believe you're drawing a meaningless distinction.

Lynn

_______________________________________________
r6rs-discuss mailing list
[email protected]
http://lists.r6rs.org/cgi-bin/mailman/listinfo/r6rs-discuss

Reply via email to