bear <xxxxxx@sonic.net> writes:

> It could be done that way, I guess, provided that
>
> (eq? sym (string->symbol(symbol->string sym))) -> #t
>
> for all symbols is preserved.

Yes, that's a requirement.

> But honestly, I don't want to
> specify case-folding in a character set that makes case-folding
> as hairy as Unicode does.

As I said, the standard should neither require nor prohibit
case-folding.