Well, I think, it might be easier to change the dmd implemention to use C only and then write language bindings to that. We all know the binding situation to C++ won't change sooner or later.
> "Daniel Murphy" <yebbl...@nospamgmail.com> wrote in message > news:jgj9mu$1q60$1...@digitalmars.com... >> On a related note, how much interest is there around here in having an >> official version of dmd written in D? >> > > I'm interested in a D *API* for taking in D sources and spitting out the > user's choice of either the parser results, or an AST with all the > semantics/CTFE/etc already run. I get the impressiona lot of people are > intrested in this. > > As far as the actual *implementation* behind the D interface, I don't > particularly care if it's C, C++, or D. > > I suspect having it D might be a pain until a lot more issues get > resolved. A bootstrapping compiler, I would imagine, would need a much > more stable base than other types of software would need (though I don't > have any experience with bootstrapping compilers, so I could be wrong).
