[R6RS] issues in document/lex.tex

William D Clinger will at ccs.neu.edu
Sun Aug 27 09:16:57 EDT 2006

Mike wrote:
> I'm not tied to the word "S-expression"---what was important to me was
> to distinguish between the <datum> and the value that it represents.
> The R5RS doesn't make that distinction.

Searching the R5RS, it appears to me that <datum>
always means an external representation, except for
some fairly harmless abuse of the term in the
descriptions of case expressions and the syntax of

> I preferred "S-expression" as
> "external representation" makes an implicit reference to the object
> represented by the piece of syntax.  (And it's harder to recognize as
> an actual term.)

In the US Lisp community, S-expression often
means the object itself, not the syntax of its
external representation.  (That meaning came
about in part through the influence of Interlisp,
whose primary representation of programs was as
lists in memory, not as characters in files.)
Given the nature of your concern, "S-expression"
is definitely a step in the wrong direction.

>  Maybe "form" would be acceptable?

Mmmm.  You're the project editor, but I think
"external representation" would be better.

> Do we leave -nan.0 unspecified then or do we specify it to be the
> reslt of (/ -0.0 0.0) ?

We should leave its meaning unspecified, because
its meaning isn't important enough for us to worry
about right now.  I just don't want its meaning to
be specified as the wrong thing.


More information about the R6RS mailing list