Re: [stack] Why is point-free form so interesting?

John Carter <[email protected]> wrote:
> William Tanksley, Jr wrote:
> >> Parameters probably should be a set, and a binding should be a map (an
> >> unordered set of pairs).
> > You're in the wrong discussion group -- concatenative languages don't
> > have parameters as such :-).

> The essence of the point under discussion is that this _is_ the
> Right Group....

Yes, and furthermore, I meant that as a light toss-off joke. I also
intended to discuss the issue further, and forgot to. My mistake; I
made it look like I was dismissing your point. In fact, I did -- and
that wasn't my intention.

> The point under discussion is representing parameter passing as stack
> operations creates a burden that seems superfluous.

Correct.

> The question is can we retain functional composition is represented by
> string concatenation whilst choosing a logically simpler (but
> implementationwise more complex) representation of parameter passing?

Stevan's method (classical visual programming) works... But is rarely used.

http://en.wikipedia.org/wiki/Visual_programming_language

Named parameters lose dataflow -- or at least they CAN. Maybe there
are rules to prevent that?

Array dataflow languages are perhaps our best bet. But not any of the
existing ones, I think. I speculate that we want a stack language with
a very powerful array-building (or multiset-building) sublanguage, so
that independent dataflows can be elegantly gathered together into a
single stack element.

Okay, that's wild speculation. I really don't know. Some experiments
are taking shape in my head... I'll have to write these things down
and see what comes of them.

> John Carter                             Phone : (64)(3) 358 6639

-Billy