[Axiom-developer] Re: bootstrap Meta/Boot

[M. Edward (Ed) Borasky]

address@hidden wrote:
>> the one that I think has the strongest legs is Erlang
> 
> Funny you should mention that as I'm half way thru the new Erlang book.
> This model, if not the language itself, is a perfect vehicle for an
> implementation of provisos (much discussed on this list). Using closures,
> SBCL threads, and some transparent networking I believe this could all
> be implemented in common lisp. Laying this model under the algebra gives
> us the ability to write parallel algebra using provisos transparently.
> 
> Tim

Portland Oregon hosted last year's functional languages conference, and
I went to two of the language-specific workshops on the weekend before
the conference itself. It was an agonizing decision -- all of the major
functional languages were represented, and the only one I really
consider myself fluent in is a subset of Common Lisp. In the end,
though, I chose Erlang and Scheme.

For what it's worth, I think the Schemers were right in the great
Lisp/Scheme schism, and having spent nearly 3/4 of my career writing
assembly language and FORTRAN for a living, I have a great appreciation
for the power of low-level languages. They're addicting -- perhaps the
most addicting of all is Forth.

Also, the Scheme workshop featured a fascinating paper by the Gambit-C
Scheme team on Termite, where they've implemented the primitives of
Erlang concurrency in Scheme with competitive performance on a number of
tests with Erlang itself. Presumably it could be implemented in SBCL if
it can be implemented in Scheme. Google for Scheme Gambit Termite and
you'll find it.