On 12/3/11 10:53 AM, Florian Weimer wrote:
Is this to related to Weirich et al., "Generative Type Abstraction and
Type-level Computation", in the sense that it is a precursor of the
problems discussed there?
I hadn't seen that paper before. I read through it. I don't know how
related the problems they describe are to the problem that I describe.
They mention that "newtype" is often used in Haskell as a way to
workaround the absence of multiple implementations for a single
interface. Using a scoped system like we've been discussing obviates
the need for this workaround, which is good.
Niko
_______________________________________________
Rust-dev mailing list
[email protected]
https://mail.mozilla.org/listinfo/rust-dev
