On Wed, Nov 3, 2010 at 8:15 PM, Ben Karel <[email protected]> wrote:

>
> From what I've read of the literature so far, the most well-known
> approaches are regions, ownership, and separation logic.
>

Of these, I think only regions provide a complete solution. I'll return to
this momentarily.


>  Disciple, for example, tracks laziness as well as mutability via type
> class constraints, and provides syntactic sugar to approximate using mutable
> as a type qualifier. The notion of spine-lazy vs element-lazy lists is
> exactly analogous to distinguishing between a mutable list of immutable
> elements and an immutable list of mutable elements.
>
> DPJ allows...
>

Thanks for those pointers. I'll have a look at both.
_______________________________________________
bitc-dev mailing list
[email protected]
http://www.coyotos.org/mailman/listinfo/bitc-dev

Reply via email to