Updating the IEEE Scheme standard

[Christopher Haynes <chaynes@cs.indiana.edu>]

I received five direct replies to my note on the need to reaffirm or revise
the IEEE Scheme standard.  Three recommended reaffirmation and the other two
very minor modifications.  Others I talked with at the Lisp conference
expressed similar feelings.

My note mostly dealt with revision because I was led to believe (by two
IEEE standards staff members) that reaffirmation only requires such a
recommendation by the working group, whereas revision is more complicated.
I have since learned what seems to be the real story from "IEEE Staff
Engineer" Kathy Doty.  (Pardon me for not digging everything out of the IEEE
standards manuals!)

Reaffirmation turns out to require essentially the same formal ballotting
process as revision, including distribution of the old standard to the
ballotting group.  To start the reaffirmation process the working group
simply informs its sponsor (the MSC in our case) that the standard contains
no known erroneous or out of date information.

My note also indicated, based on erroneous advice from IEEE staff, that the
time required between submission of a draft and its approval is about six
months.  In fact, it is more like a year, or more if resolution of negative
ballots is not trivial.  The good news is that the standards board will not
administratively withdraw the standard in December 1995 if administrative
steps have been taken to revise or reaffirm the standard.  However, if a
revision is desired but will not be completed close to that date, the
appropriate action is to exercise yet another option: requesting that the
standard be "continued" for up to two years, which requires 75% of at least
75 votes in a formal ballot.  A reaffirmed standard, like a new or revised
one, is good for five years, but may be revised sooner (generally after
two years or more).

With this added information the majority may still prefer to simply
reaffirm the standard.  (That would be fine with me.)  However, the new
information is that as far as the IEEE is concerned this is
administratively not much different than revision.  Thus we might want to
make a few changes, perhaps so we can honestly say it contains no known
erroneous or obsolete elements.  It seems clear, however, that no one wants
an open-ended process with attendant worries about where it will end.  

I suggest that we allow a reasonable period, say one month, in which
specific revision proposals may be presented in this newsgroup, and that a
vote of the newsgroup members then be taken to either reaffirm the standard
or draft a PAR for revision that will specifically limit consideration to
those proposals for which there is substantial support.


PS: I thought I'd posted this more than a week ago, but we had mail problems.