Re: What happened to the ex-Guile VM?

[Marius Vollmer]

Keisuke Nishida <address@hidden> writes:

> At 20 Mar 2001 22:52:34 +0100,
> Marius Vollmer wrote:
> > 
> > > I found somebody was working on a bytecode interpreter (VM) for
> > > Guile around 1997.
> > 
> > Was this the coupling of Guile to a Java Virtual Machine, "Latte"?
> 
> I'm not sure.  This message:
> 
>   http://www.red-bean.com/guile/guile/old/2249.html

I see.  This is not "Latte", but I think I rembember reading about it.
(But I wasn't really involved in Guile at the time).

> > It will also be important to know how stanards conforming your VM is.
> > Does it handle continuations correctly?
> 
> Within the VM, yes, by stack copying.  But I haven't thought of
> C level continuations.  Should they be supported?

What is a C level continuation?  I think we should suppport
continuations like we support them now: intermediate calls to C
functions are part of the continuation and the C stack needs to be
copied as well.

Since C has no defined semantics for continuations, we can define our
own.  I'd say it is OK to just specify that continuations are
implemented by copying the stack (or appear to be implemented this
way).  This has consequences for modifications to local variables
since these avariables become sahred when the stack is copied.