John Cowan <xxxxxx@ccil.org> writes:
> Another is to write your own functions that always use the default
> comparator

The problem with that is that these short&sweet names are spoken for in
a 'final' SRFI on the track for a RnRS in some capability or other.

(To the extent that the present semantics are carved in granite, I was
mislead when requested to participate here.)

The problem with the dispatching versions isn't when comparing
comparators, it's when comparing unknown objects which might or mightn't
be comparators.

But it could be (=? "hi" "hi") and (=?/with-comparator d "hi" "hi") or
=?* or another name of your choosing for the version that needs a
comparator arg.

As in, instead of dispatching, they could use the registered comparator
always. That is the entire point of having registration.

> as they are one-liners.

Of course. This is a design topic more than an implementation problem.