Aubrey Jaffer <xxxxxx@alum.mit.edu> writes:

> In R5RS the eof-object has no read-syntax. If it did, then calls to
> READ would be unable to distinguish between an actual end of file
> and an eof object written in a file.

This reason does not apply to NaN. It's not used to signal the end
of file.

> Okay, how about a library function:
>
> (define NaN (lambda()

What advantage it has over a literal syntax? A NaN must have a literal
syntax anyway for output. So why not use it for input too? All Scheme
values which have a defined output syntax use it also for input.

> So your code is clear and readable, and yet you don't have a
> read/write syntax that makes it unclear whether read is failing or
> just reading an error.

Read should never fail on NaN. If it's present in the stream provided
for reading, it's meant to be read.

--
   __("<         Marcin Kowalczyk
   \__/       xxxxxx@knm.org.pl
    ^^     http://qrnik.knm.org.pl/~qrczak/