[Zlib-devel] zlib 1.2.2.3 available for testing: OSF/1 4.0 Alpha failure

[Nelson H. F. Beebe]

The proposed simple patch for gzio.c

% diff -u -r gzio.c~ gzio.c
--- gzio.c~     2005-05-30 17:13:36.000000000 -0600
+++ gzio.c      2005-05-30 17:54:50.000000000 -0600
@@ -635,7 +635,7 @@
                        a11, a12, a13, a14, a15, a16, a17, a18, a19, a20)
     gzFile file;
     const char *format;
-    int a1, a2, a3, a4, a5, a6, a7, a8, a9, a10,
+    long a1, a2, a3, a4, a5, a6, a7, a8, a9, a10,
         a11, a12, a13, a14, a15, a16, a17, a18, a19, a20;
 {
     char buf[Z_PRINTF_BUFSIZE];

did indeed fix the failures on  OSF/1 4.0 Alpha.  

I did four different builds with cc and gcc on that system with the
patch applied, and all four reported the expected 

	"*** zlib test OK ***".

Now, as to the question of whether we can abandon <varargs.h> in favor
of the 1989 ISO Standard C's <stdarg.h>.  Here is some evidence: every
one of the 20+ different Unix platforms that I routinely test software
on has <stdarg.h>, and recent gcc releases refuse to compile code with
<varargs.h>.  Even the C implementation on the PDP-10 running TOPS-20
has a <stdarg.h> (dated 11-Aug-1989): this runs in a PDP-10 emulator,
since our DECsystem-20/60 hardware was retired in 1990.

-------------------------------------------------------------------------------
- Nelson H. F. Beebe                    Tel: +1 801 581 5254                  -
- University of Utah                    FAX: +1 801 581 4148                  -
- Department of Mathematics, 110 LCB    Internet e-mail: beebe at math.utah.edu  -
- 155 S 1400 E RM 233                       beebe at acm.org  beebe at computer.org -
- Salt Lake City, UT 84112-0090, USA    URL: http://www.math.utah.edu/~beebe  -
-------------------------------------------------------------------------------