RE: [colorforth] Machineforth IFs

[Fréderic DUBOIS <frederic.dubois@xxxxxxxxxxxxx>]

> -----Message d'origine-----
> De : Mark Slicker [mailto:maslicke@xxxxxxxxxxx]
> Envoyé : mar. 4 février 2003 21:02
> À : colorforth@xxxxxxxxxxxxxxxxxx
> Objet : RE: [colorforth] Machineforth IFs
> 
> 
>In a sense they do. The set of objects in this case are the integers
>modulo 2^32, with the addition and multilpication operators they form a
>commutative ring. 'jump' and 'if' don't seem to fit in the algebra because
>they are not purely operations on the set but effect the flow of control.

That was my point. The idea is to find the ring to eliminate any kind of
control-flow.
A program then would be a set of interleaved rings. However programming in
such an environment has
drawbacks because some real-life aspects may not model very conveniently.
But maybe I'll explore
for myself this idea of 'finding the rings' as a design method and a way to
think through the problem,
as someone said ;).

>> On c.l.f someone posted a link to an article dealing with linear
>> programming
>I think you refer to this document:

>http://home.pipeline.com/~hbaker1/ForthStack.html

>This also inrigued my interest, I'd like to learn more about this linear
>logic when I get a chance.

Yes it is. I should moderate my enthousiasm however since ROLL seems to be
primordial there.

>Mark

---------------------------------------------------------------------
To unsubscribe, e-mail: colorforth-unsubscribe@xxxxxxxxxxxxxxxxxx
For additional commands, e-mail: colorforth-help@xxxxxxxxxxxxxxxxxx
Main web page - http://www.colorforth.com