Re: views of binary operations

erk wrote:
>
> At the risk of repeating what you already know, I think Marshall
> intends the unadorned assignment (r = A op B) to assign a value
> (snapshot) to r. But the view doesn't assign a value to r - it's a lazy
> evaluation (but without the closure that usually goes along with it).
> The hybrids are partial lazy evaluations - as such, I guess they're
> closures. For example, "r = view(A) op B" closes over B, capturing its
> value at the time this statement is evaluated, but only "capturing" A's
> value when the value of r is used. So a temporary, local snapshot is
> created.

Yes, that was my idea. The question is, is it useful to be able to slice things up that finely? I kind of think not, but I'm not sure.  

Marshall Received on Tue Jul 18 2006 - 07:10:46 CEST