Shriram Krishnamurthi <xxxxxx@cs.rice.edu> writes:

> Per Bothner wrote:
>
> > One reason for using a single name is that I'm interested in
> > experimenting with alternative syntaxes, including use of infix
> > operators. [...]
>
> I fail to see why a new operator, GENERALIZED-MUTATE (say), could not
> meet your needs just as well.  ... Why does the chosen name have to be SET!
> and nothing else?

It doesn't.  I was explaining where I was coming from, and one reason
I feel it is natural to have a conflated mutator, at least in languages
that use an infix syntax.  Most of the arguments against generalized-set!
seem to have been against the general idea of a conflated mutator,
even their use in languages that do use infix syntax.

> I just want to know why that one name, which already has a fixed
> syntax and semantics in standard Scheme.

Perhaps because I've been using Kawa as a framework to work out some
ideas on language design, in the context of a language that is
compatible with Scheme.  That approach may be fine for Kawa; perhaps
that is not appropriate for Scheme in general.  I do admit in some
ways Kawa is straying from "the spirit of Scheme": for example it has
optional type specifiers.  (Worse, use of types is quite ad hoc, at
this time.)

Still, there are at least two Scheme dialects that *do* implement
extended set!, so it seemed to make sense to make a srfi for it.

> After all, if your source language has := rather than SET!
> (which would likely be a poor choice of name in a traditional infix
> syntax, since `!' may well mean something else), then your translator
> can pick any old name it wishes in the target language.

Yes.  I wrote something similar, but edited it out.
--
	--Per Bothner
xxxxxx@bothner.com   http://www.bothner.com/~per/