A Guide to the R7RS Steering Committee Candidates

Will Clinger posted “a politically incorrect guide to the candidates” in hopes of helping registered voters here.

It is well written and worth a read by anyone interested in the future of Scheme.

The following is a copy of the entire message:


The Scheme Language Steering Committee's announcement
of the forthcoming election says

    When the nomination period ends, the complete list
    of candidates will be published on www.r6rs.org.

    Candidates may also post whatever messages they wish
    to comp.lang.scheme, the r6rs-discuss mailing list,
    or whatever other forums they feel appropriate, and
    voters should feel free to discuss the candidates
    and their positions on these fora.

After 12 candidates have been nominated, and 129 voters
registered, we now begin the campaign and/or endorsement
period.

Voters will be asked to rank the candidates, and three
candidates will be elected by "single transferable vote
proportional representation."

The top two votes on my ballot will be John Cowan and
Jonathan Rees (in some order).  My reasoning is simple:
In my opinion, any subset of the twelve candidates that
includes those two would make a fine Scheme Language
Steering Committee.

The Steering Committee is responsible for the process
and its direction.  As Rees and Cowan indicated in their
statements, they have experience with process issues and
also understand that the direction of this process needs
to change.

That does not make them unique among this group of twelve
candidates.  Indeed, I have not decided which of several
worthy candidates I should rank third.

The candidates can be aggregated along several dimensions.
Three voted to ratify the R6RS; five voted against
ratification; four did not vote.  Three candidates were
editors of the R6RS, and two more served on the editors'
committee but resigned before any documents were put to
a vote.  Eight candidates have implemented or are now
maintaining a popular implementation of Scheme.

There is much to be said for the neutrality that comes of
not being associated with any particular implementation;
all three members of the original Steering Committee were
neutral in that sense.  Of the twelve new candidates,
four are not associated with any major implementation:
John Cowan, Anton van Straaten, Ray Dillinger, and
Richard O'Keefe.

John Cowan is an expert on Unicode, and his comments
on drafts of the R6RS showed excellent judgment.

Anton van Straaten was the only R6RS editor who was not
associated with a particular implementation.  That gave
him a broader perspective that was sometimes difficult
for the other editors to appreciate.  He was the only
one of the four R6RS editors who abstained from the
vote on ratification.

As Ray Dillinger noted in his statement, he took the
initiative to renew IEEE Standard 1178, which is still
the only standard for Scheme that is recognized as a
national or international standard.  Dillinger did not
vote on R6RS ratification.

Richard O'Keefe is an accomplished Scheme programmer.
Many implementations of Scheme rely on his efficient
code for merge sorting of lists.  O'Keefe did not vote
on R6RS ratification.

Turning to the implementors, Will Clinger is the only
candidate who has implemented the R6RS.  Larceny was,
in fact, the first substantially complete implementation
of the R6RS.  If you think that's a good reason *not*
to vote for Clinger, then I have no argument with you.

If you'd like to vote for an implementor who has shown
less enthusiasm for the R6RS than Clinger, then you
have seven to choose from.

Marc Feeley polled implementors of the R3RS/R4RS/R5RS
and IEEE/ANSI Scheme to gauge their enthusiasm for the
draft R6RS that was put up for ratification; that was
something the editors themselves should have done.
Marc also served as the original chair of the R6RS
editors' committee.

Aubrey Jaffer, the implementor of SCM, has also been
the driving force behind SLIB, which was arguably the
first successful collection of portable libraries for
Scheme.

Chris Hanson, who has been maintaining MIT Scheme, is
a charter author of the R*RS documents, and was an
editor of the IEEE-1178 standard.  He also wrote much
of that standard's Appendixes B and C, which were
significant milestones during the standardization of
Scheme (and Lisp generally).

Jonathan Rees was one of the original implementors of
T and Scheme 48.  He too is a charter author of the
R*RS documents, and was editor of the much-beloved
R3RS.

Olin Shivers is responsible for scsh.  Among the
twelve candidates, he is the only implementor who
abstained from the R6RS ratification vote.

Kent Dybvig is responsible for Chez Scheme, and did
a good job of chairing the R6RS editors' committee
after Feeley resigned.  In his statement, Kent Dybvig
said he "will not use a position on the steering
committee as a mechanism to push" his opinions.

Mike Sperber, who has been maintaining Scheme 48,
edited the R6RS documents.  He has also served as an
editor for the SRFI process, which was arguably the
second successful collection of portable libraries
for Scheme.

In October 2007, Dybvig and Sperber announced their
intentions to implement the R6RS in Chez Scheme and
Scheme 48 (respectively) within a year.  There is
much to be said for their moderate approach to
implementing the R6RS, just as there is much that
could be said against the pioneering approach taken
by Ikarus, Larceny, PLT Scheme, Ypsilon, IronScheme,
and Mosh.

I wrote this guide to the candidates in hope it will
help some voters.

Someone nominated me.  As the only candidate who is
responsible for maintaining implementations of both
the R5RS and R6RS, I certainly have a stake in the
outcome of this process.  If elected, I will serve
to the best of my ability.  Urging you to vote for
me would have been the politically correct thing for
me to do.  Instead, I urge you to vote for the best
committee you can imagine.

Will

Leave a Reply

Your email address will not be published. Required fields are marked *