Re: [stack] disallowing recursive definitions

On 2/4/08 1:49 PM, "John Cowan" <[email protected]> wrote:

> Manfred Von Thun scripsit:
> ....
> Indeed.  But I was only saying that arguments to a word are nothing
> *fundamentally* new, unlike (say) alternation.  That doesn't mean that
> the syntactic sugar isn't worth having.
> 
What a coincidence. Last weekend I started work on a non-deterministic
postfix stack language, which of course has alternation as one of its
principal program forming operators. That makes it non-concatenative
I suppose.
> 
> ...
> Modifiers to primitives may not be that useful, but modifiers to
> user-written words are a different story.
> 
>> > If anything like this is introduced, then a decision has to be made
>> > whether modifiers are just for the primitives or also for user defined
>> > molecules.
> 
> I think it's important to have them for both.

To have both would imply a huge revision of the language. To have modifiers
just for primitives would be much easier. And there is a further, more
modest, choice: Find a collection of fixed modifiers that make sense to
apply to (just about) anything. Again I cannot think of any such beasts
whose actions cannot be achieved equally well by combinators. They would
be variants to the i combinator:

[...] i
[...] [f g h] flagged-i

Now it may be that there are no flags of any kind that whose effects cannot
be achieved by existing mechanisms. Well, that may be failure of the
imagination,
or it indicates that the existing machinery is really quite good already.

 - Manfred




[Non-text portions of this message have been removed]