Re: [stack] Cat article submission to Doctor Dobbs Journal

Stevan Apter <[email protected]> wrote:
>  From: "Christopher Diggins" <[email protected]>
>  >> > You keep saying concatenative languages, and I am talking about
>  >> > stack-based languages. I am still not sure which languages are
>  >> > technically concatenative (is Forth? is JVML? is PostScript?).

>  >> i've been up and down this road so many times i'm giving names to
>  >> the rocks. am i wrong in thinking that the languages we're
>  >> concerned with (in this group) are those in which wholes are built
>  >> from parts by concatenation and where concatenation denotes a
>  >> fundamental operation, either composition or application (or
>  >> something else)?

>  >> joy and its descendents satisfy that definition. other languages
>  >> can be restricted in various ways to satisfy it.

>  > This seems more or less like the definition that majority of the group
>  > has agreed upon. In your opinion what languages other than Joy does it
>  > include?

It's close. I think our last discussion fruitfully concluded that a
concatenative language:

1. Syntactically is concatenated from a set of primitives;
2. Semantically maps concatenation to an associative operation.

I'm not certain whether we MUST make function composition the only
associative operation; I'm also not certain that it's strictly
necessary to specify the semantic mapping (because concatenation is
itself associative), but I like having it explicit, because it makes
it clear that the semantics and the syntax have the same property
(associativity). There's a beautiful symmetry there... Beauty is
truth, and truth beauty.

Application doesn't work because it's not associative.

>  just a guess -- billy will correct me if i'm wrong:  the tacit part
>  of J (application), point-free haskell, one (or more?) of jot/zot/iota,
>  chris okasaki's flattened combinator language, lee spector's PUSH,
>  van Oormerssen's False.

Most of these don't meet that definition, I think. I still haven't
figured out jot, zot, and iota... I'd like to know whether they're
concatenative.

>  i'm still not sure in what sense or under
>  what restrictions Forth or Postscript satisfy the definition.  it
>  would be good to be able to answer the question "is this language
>  concatenative, and if not, why not?"

I think the recent consensus gives a clear test. I can't answer for
jot etc because I can't read their notation.

-Wm