std-unix@longway.TIC.COM (Moderator, John S. Quarterman) (08/23/89)
An Update on UNIX* and C Standards Activities
August 1989
Jeffrey S. Haemer, Report Editor
USENIX Standards Watchdog Committee
ANSI X3J11 C Language Update
Doug Gwyn (gwyn@brl.mil) reports:
There's not much new on the X3J11 (ANSI C) front.
As of about a week ago [i.e, mid-May, 1989 - jsh], X3 had not yet
finished the reballoting caused by having to respond to a previously
lost, public comment letter from Russell Hansberry. X3J11 discussed
these comments with Hansberry at the Seattle meeting, voted on some
resulting proposals, and, in summary, reaffirmed previous resolutions
of and decisions about all his issues. In all, no changes were made
to the December 1988 draft proposed standard and rationale documents.
An official response was sent to Hansberry, who had 15 working days to
respond to X3, after which X3 would again ballot on whether or not to
send the proposed C standard to ANSI for ratification. Hansberry
replied, requesting a full formal review process. Since this was
previously approved, we expect the same outcome for the reballot, but
the people involved in the appeals process are not the same as the
ones with technical expertise who drew up the standard, so anything
could happen. Certainly there will, at least, be a substantial delay
in obtaining final approval of the submitted standard as an ANSI
standard.
ISO WG14 met concurrently in Seattle. A Danish proposal for an
alternative to trigraphs was defeated by both X3J11 and WG14; although
one might hope that we've heard the last about this, the delay on the
ANSI side might permit more hassle from the Danes. WG14 also agreed to
submit the same proposed standard as ANSI's for ISO approval, with the
understanding that British concerns about excessive instances of
"undefined" behavior would be addressed early in the X3J11
"interpretations" phase. Specifically, the British would like all such
instances clearly identified. X3J11 is working with them to prepare
an "information bulletin", which would clarify the standard without
forcing a revision of the proposed standard itself.
X3J11 work for the foreseeable future will concentrate on answering
__________
* UNIX is a registered trademark of AT&T in the U.S. and other
countries.
Jeffrey S. Haemer, Editor USENIX Standards Watchdog Committee
August 1989 Standards Update - 2 - ANSI X3J11 C Language
questions about the standard and providing rulings on interpretations.
No new instances of X3.159/1003.1 conflict have arisen, to my
knowledge, since the "great `environ' problem". There have been
several varying interpretations of how vendors should define __STDC__
(if at all) in an "extended" implementation of X3.159, such as most
POSIX vendors will be doing for reasons of backward compatibility.
X3J11 certainly intended all positive integral values of __STDC__ to
be reserved for strictly standard- conforming implementations of C;
there is some disagreement whether non-positive values should be used
by vendors to indicate "ANSI C except with extensions". Unfortunately
there is no way to constrain non-conforming implementations via
wording in the standard.
A proposal that X3J11 undertake standardization of C++ was rejected;
although there was a consensus that C++ was ready for a standards
effort to begin, it was not felt that C++ should be undertaken by
X3J11 itself, for a variety of reasons.
Rex Jaeschke has formed a "Numerical C Extension Group", which has
begun work on identifying extensions needed for C to fully serve the
numerical computing community. This is not yet officially under X3
auspices, but it could become so.
The X3J11 meeting slated for September, 1989 in Salt Lake City was
canceled due to the approval delay; the next scheduled meeting is in
New York City on March 5-6, 1990.
Jeffrey S. Haemer, Editor USENIX Standards Watchdog Committee