Re: mixing lush with clush

On Thu, 16 Mar 2006, Yann LeCun wrote:

>
>> Sorry, I'm not clear on the differences between DX, DY, and DH.
>> What does it mean to generate DXs instead of DHs?
>
> DX and DYs are are interpreter functions written in C.
> A DH is a function generated by the Lush compiler.
>
> Contrary to Ralph, I would make everything more like
> the DH and progressively remove DX/DY and replace
> them by a minimal interpreter (everything would be
> compiled on the fly).

Yes! Our ideas might actually be more similar than they sound.

I was suggesting DXs\DYs instead of DHs because they are simple
enough so that I can understand them. The "type-info extensions"
that I was dreaming of would probably reproduce much of what's
in the DH/DHDOC machinery today minus some fancy stuff for escape
analysis and what not. But again the DH stuff is very complicated
and it might be easier to start from scratch.

Another thing that I think is wrong with the DH compiler is the
way the compilable functions are currently coded (you find them
compiler/dh-macro). It's unreadable, hard to maintain code.


>> At some point or other you're going to have to talk to the
>> interpreter, right? Even if you're okay with having to compile all of
>> your code, in order to support something lambda (or globals) you'd
>> still need to be able to encode the lexical environment.
>
> Indeed, but that can be pretty minimal.

I am not a compiler person and for a start I had a simple,
naive approach in mind in which the interpreter serves as a
convenient runtime system that provides memory management and
evaluates lambdas for me.

Ralf


-------------------------------------------------------
This SF.Net email is sponsored by xPML, a groundbreaking scripting language
that extends applications into web and mobile media. Attend the live webcast
and join the prime developer group breaking into this new coding territory!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642