[PRL] random flame: actors versus OO

[Mitchell Wand]

> This reminds me: the other day Felix and I were wondering about the 
> connection between actors and the history of Scheme. I've heard that 
> the early development of Scheme had something to do with people playing 
> with actor semantics. Aside from the general fact that Scheme is a good 
> for writing interpreters and playing with language ideas, I'm not 
> really seeing any particular relation with actors. In fact, FWICT, 
> there's nothing directly addressing concurrency in the core language at 
> all. What's the historical connection?

As a matter of fact, Scheme grew out of Gerry and Guy sitting down and
trying to make sense out of Hewitt's actors papers.  They finally
figured out that Carl's notion of "acquaintances" was just lexical
scoping, and that his notion of <I forget his word here> was exactly
cps. 

I believe that very early versions of Scheme had both closures and
actors, until Guy and Gerry noticed that they were isomorphic.

And the first Scheme Report had some concurrency primitives.  You
should go read it.

--Mitch