v05i005: xwps3, Patch1

[Dan Heller]

Submitted-by: uunet!cs.buffalo.edu!sigmast!planck!meyer
Posting-number: Volume 5, Issue 5
Archive-name: xwps3/patch1
Patch-To: xwps3: Volume 5, Issue 2

[ From the Author: ]

I have been informed of a major problem with my 'xwps' program (which
has been dubbed 'xwps3' by our moderator) which will cause it to
fill filesystems with NULLs on little endian machines (i.e. VAXEN and
VAXStations).  It seems that a little piece of code left over from the
translation from 'xwd' causes major problems.

The original code (xwd) had to translate to big endian form on little
endian machines in order to generate a common bitmap format file.  It
doesn't even get invoked on big endian machines so I never questioned
it.  Here's the context diff.

*** xwps.old	Fri Sep 22 17:17:07 1989
--- xwps.c	Fri Sep 22 17:17:25 1989
***************
*** 144,150 ****
  Window window;
  FILE *out;
  {
- 	unsigned long swaptest = 1;
  	XColor *colors;
  	unsigned buffer_size;
  	int win_name_size;
--- 144,149 ----
***************
*** 257,270 ****
  		header.window_y = (xwdval) win_info.y;
  	}
  	header.window_bdrwidth = (xwdval) win_info.border_width;
- 
- 	if (*(char *) &swaptest) {
- 		_swaplong((char *) &header, sizeof(header));
- 		for (i = 0; i < ncolors; i++) {
- 			_swaplong((char *) &colors[i].pixel, sizeof(long));
- 			_swapshort((char *) &colors[i].red, 3 * sizeof(short));
- 		}
- 	}
  
  	openImage(out, header.pixmap_width, header.pixmap_height);
  	writeImage(out,image,header.pixmap_width, header.pixmap_height,colors,ncolors);
--- 256,261 ----
-- 
   Thinking quickly, the IBM System Jock     # Bob Meyer
uttered an incantation in EBCDIC and made    # Calspan Advanced Tech. Center
the sign of the Terminated Fork.             # ...!sunybcs!planck!meyer
   The UNIX Guru only smiled and trapped him in a recursive SED script.