Here is an updated issues list:

1) R6RS provides define-enumeration to help set up enum-types. Is
this worth having? Possible syntax is: (define-enumeration <type-name>
(<symbol> ...) <constructor>)

This would bind <type-name> to the enum-type, and constructor to a
curried version of make-enum-set that already knows what type to use.

2) Should there be a mechanism to convert between integer sets and
integers as bitvectors, as defined in SRFI 33, SRFI 60, and R6RS?

3) (resolved)

4) How about `set-intern!`, which is like `set-value` but adds the
element to the set if it's not already there?

5) Should there be a lexical syntax for sets?  If so, how do
we incorporate the equivalence procedure, which has no printed
representation?  Perhaps we only need to support the big five, or
perhaps not even `eq?` as it is implementation-dependent.

6) Are bags really useful enough to include?

7) Should bags be called multisets (msets?) instead?

8) Should we switch to unique enum objects rather than symbols?

9) The word "set" is heavily overloaded.  Should we use "gset" or
"generic-set"?

--
John Cowan            http://www.ccil.org/~cowan     xxxxxx@ccil.org
                if if = then then then = else else else = if;