Various posters have quite reasonably asked for a more precise
specification and sample implementation (the latter which
Lars Thomas Hansen has kindly provided).  Perhaps I put forth
a flawed proposal.  However, my idea was that there were various
alternative that needed to be nailed down, and I wanted to
get a consensus first.  However, this has back-fired in that the
lack for formality made people misunderstand the proposal.
Now, we have sample implementation, and hopefully the misunderstandings
have been cleared up, and we need to resolve some of these issues.

First, does anyone want to speak up on Alternative 1 vs Alternative 2?
My current inclination is Alternative 1, and unless there is ground-swell
of support for Alternative 2, the final SRFI will use Alternative 1.

Did I miss any standard procedures that should have setters?
Any procedures in existing SRFIs that should have setters?
Consider for example srfi-4.  We probably want
        (setter TAGvector-ref) -> TAGvector-set!
Should such a requirement be part of srfi-4 or srfi-17?
Adding a new srfi seems excessive.  Perhaps we can add a
requirement that any implementation that claims to
implement both srfi-4 and srfi-17 should define:
        (setter TAGvector-ref) -> TAGvector-set!
Could we add an addendum to srfi-4 stating this?

Any comments besides Lars's on the efficiency issues?
--
	--Per Bothner
xxxxxx@bothner.com   http://www.bothner.com/~per/