Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!yale!hsdndev!cmcl2!kramden.acf.nyu.edu!brnstnd
From: brnstnd@kramden.acf.nyu.edu (Dan Bernstein)
Newsgroups: comp.lang.misc
Subject: Re: joke is on you.
Message-ID: <4603:Jan421:58:1591@kramden.acf.nyu.edu>
Date: 4 Jan 91 21:58:15 GMT
References: <40569@nigel.ee.udel.edu> <3340:Jan322:21:4791@kramden.acf.nyu.edu> <19771@yunexus.YorkU.CA>
Organization: IR
Lines: 32

In article <19771@yunexus.YorkU.CA> oz@yunexus.yorku.ca (Ozan Yigit) writes:
> Not at all, but by your amusing definition,

Do you even have a definition?

> I would say ``8088 assembler
> supports Forth''

It does not, by my definition of ``supports'' (``has''). See below.

> along with arbitrary-precision arithmetic, composable
> functions, ad nauseam...

The definition of ``has'' that I posted only applies to features of
language power. Example: C has mod-2^37 arithmetic. To make interesting
statements with ``has'' you need to add more information: C on this
machine has mod-2^16 arithmetic implemented with single instructions.
This means that there exists a method, X, for each of the operations of
mod-2^16 arithmetic. X can be implemented in C on this machine. And X
turns into single instructions under C on this machine. (Of course, it's
more useful for the reader if X is spelled out explicitly.)

Question for the doubters: How do you propose to say ``C on this machine
has mod-2^16 arithmetic implemented with single instructions'' without
using a word equivalent to my ``has''?

Forth is more than semantics. Forth is (at least) syntax plus semantics.
Therefore the statement ``8088 assembler has Forth'' is nonsensical
under my definition of ``has.'' It might have occurred to you before you
posted that the word ``has'' is overloaded.

---Dan