C standards and #include

[frew at ucsbcsl.UUCP]

Our site is new to USENET, so "forgive me in advance" (so to speak) if
the following comments are totally out of sync with the current
discussion.

First, a request:  I have been following with great interest the
discussion of C standards in this newsgroup.  I am delighted to hear
that ANSI is well along on a C standard.  Could one of you gurus please
post an address and contact person from which I could obtain a copy of
the current draft?  If that's too redundant for the net please mail me
direct.

Second, an observation:  when trying to move a large (25K line) C
package from UNIX to VMS, I was immediately flummoxed by a glitch in
VMS-C's implementation of #include "filename".  The BOOK of C (A Appx.
12:2) sez:

	The named file is searched for first in the directory of the
	original source file, and then in a sequence of standard
	places.

Whereas, in "Programming in VAX-11 C":

	... if, after applying the usual RMS defaults,
	no directory is known, then the directory containing the source
	file is searched for the file.

(I assume that "the usual RMS defaults" means file-specification
parsing according to Appx C of the VAX-11 RMS reference manual).  VMS
therefore has no provision (like "cc -Idir") for augmenting the
#include search path for double-quoted filenames (include-file
"libraries" are supported but the C syntax is not portable).

For a software system of any size this is a serious drawback.  Is this
issue being addressed by the C standards folks?

James Frew
Computer Systems Laboratory
University of California
Santa Barbara, CA 93106

(805) 961-2309 (machine answers "Snowmelt Project")

ucbvax!ucsbcsl!frew