The motivation makes sense to me, but I am confused by the sentence
beginning with "Now that SRFI 153...".  How does SRFI 153 help in this
regard?  Or is it just a matter of precedent/analogy?

Also, it seems that this note would disallow (or discourage) a SRFI-113
implementation that is not based on hash tables (e.g., based on balanced
trees).  Is the intention that SRFI 113 should focus on hash table
implementations (and optionally tree/ordered ones) while SRFI 153
provides tree/ordered implementations?

Regards,

-chaw

> <p><b>Post-finalization note 3</b>:
> The "Comparator restrictions" section of this SRFI states that
> implementations must not require comparators used to create
> sets or bags to have both an ordering predicate and a hash
> function.  However, in practice this is a nullity, because there
> is no way to determine which type of procedure the implementation
> requires, and therefore users must supply both procedures.  Now that
> <a href="http://srfi.schemers.org/srfi-153/srfi-153.html">SRFI 153</a>
> (in draft status at the time of this writing) requires comparators
> with ordering predicates, the author of this SRFI strongly urges
> implementers to accept comparators with hash functions with or
> without ordering predicates. The sample implementation,
> which is built on top of hash tables, already does so.</p>
>
> Comments?
>
> --
> John Cowan          http://vrici.lojban.org/~cowan        xxxxxx@ccil.org
> Time alone is real
>   the rest imaginary
> like a quaternion       --phma
>