On Mon, Oct 27, 2008 at 12:02 PM, Brian Alliet <brian@brianweb.net> wrote:
On Mon, Oct 27, 2008 at 10:58:11AM +0000, Simon Peyton-Jones wrote:
> Is there an interest in hosting GHC on the JVM (besides my own).

Yep. I wrote a JVM backend for GHC (LambdaVM). It is suffering from
bit-rot though. I think this thread has re-spaked my interest in it
though.

> I don't think it's so hard to translate GHC's Core language to the
> JVM.

Definitely not. That is more or less the route I took (I actually
transform STG it into yet another simple intermediate laguage that is
more JVM friendly, this was more for optimization purposes though).
While C-- looks fantastic for generating native code, it just didn't
seem worth the effort to shoehorn it into the JVM. Although I haven't
looked at any of the new backend stuff, I suspect it still won't be
suitable for the JVM.

-Brian
_____

Being able to integrate Haskell into a rather large Java codebase might have made me not choose Erlang.  Luckily, the OTP Erlang distribution had a lot of other really desirable features for the project I was working, yet I feel the lack of strong type checking hurt my productivity a bit, especially when trying to come up with a way to specify the real INs and OUTs between the Java and the Erlang code in terms of types.  

The right way to solve that problem is with a stronger specification between the Java team, and the Erlang team (aka me), however, it was still a really positive experience in the ways of functional programming, and I suppose I could have used Haskell to generate the glue between Java and Erlang had I been a better Haskeller :-).

 
__________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe