On 17 May 2013 19:09, John Cowan <xxxxxx@mercury.ccil.org> wrote:
> Can I assume that means everyone on this list is entirely happy with
> them exactly as written?  Ghu knows, boxes are simple, which is why
> I picked them as the first R7RS-large effort.  But a little feedback
> wouldn't hurt either.

The spec seems good.  I would personally make the lexical syntax
optional, as it is (to my eyes) ugly and doesn't provide anything
strictly necessary, which the rest of the spec does.  It might be
worth noting in the spec that the reference implementations do not
implement the lexical syntax.  I don't find any of the optional
autoboxing functionality useful, and it slightly complicates the
mental model of procedure calls, but overall it's probably appropriate
in the spirit of a large spec.

I don't really see the need to discuss the relation of boxes to
promises in the spec.

Cheers,
--
David Banks  <xxxxxx@gmail.com>