[Yhc] is YHC dead?

[Neil Mitchell]

Hi,

> I've managed to use old Brian Alliet project yhc2jvm to compile some
> hello-world haskell code to JVM.
> I'm close to having a custom JVM ClassLoader load and translate YHC's
> bytecode on the fly.

Very cool!

> I haven't looked much into YHC internals so far (I'm still a Haskell
> newbie).
> Does it do any optimizations on the code before or after yhc core is
> produced?

No, it doesn't even optimise things like "otherwise", so you see lots
of things like case otherwise of True -> real code; False -> error.

> If JavaScript is created straight from Yhc.Core than maybe I'd be also
> better to produce JVM bytecode without YHC bytecode phase...

It's your choice entirely. The pipeline is Haskell -> Core ->
bytecode. You can jump in at any level - I wouldn't be able to say
which of Core/bytecode makes a more appropriate target.

> BTW, I've tried other aproaches to executing Haskell on
> JVM: http://code.google.com/p/jhugs/ ;-)

Your work sounds cool, so I would encourage you to try and replicate
some of these results in GHC (which also has Core and a bytecode,
although in fairly different ways). GHC is a lot more work to get to
grips with, but your results will be much more widely useful -
allowing others to play with your ideas.

Thanks, Neil

> On Sun, Apr 3, 2011 at 8:50 PM, Neil Mitchell <ndmitchell at gmail.com> wrote:
>>
>> Yhc is indeed dead. I've posted a blog post
>> (http://yhc06.blogspot.com/2011/04/yhc-is-dead.html) and updated the
>> wiki (http://haskell.org/haskellwiki/Yhc).
>>
>> Thanks, Neil
>>
>> On Fri, Apr 1, 2011 at 12:53 PM, Malcolm Wallace <malcolm.wallace at me.com>
>> wrote:
>> >>   Is YHC compiler project still alive?
>> >
>> > No.  There has been no-one working on YHC for at least two years now.
>> >
>> > Regards,
>> >    Malcolm
>> >
>> > _______________________________________________
>> > Yhc mailing list
>> > Yhc at haskell.org
>> > http://www.haskell.org/mailman/listinfo/yhc
>> >
>
>