Path: utzoo!utgpu!news-server.csri.toronto.edu!clyde.concordia.ca!uunet!decwrl!adobe!delahunt
From: delahunt@adobe.COM (Jim DeLaHunt)
Newsgroups: comp.windows.ms
Subject: Re: Printing WinWord remotely? /// Win3.0 Genoa VGA driver?
Summary: %! is convention, ^D is protocol; neither is PostScript.
Message-ID: <4142@adobe.UUCP>
Date: 20 Jul 90 01:21:44 GMT
References: <4995@milton.u.washington.edu> <DSAMPSON.90Jul16111525@x102a.harris-atd.com>
Reply-To: delahunt@adobe.com (Jim DeLaHunt)
Organization: Adobe Systems Incorporated, Mountain View
Lines: 79

This discussion on %! and control-D in PostScript(R) language files 
perpetuated some common misconceptions; I'd like to clear them up.
I hope David Sampson will forgive me for picking on his message.

In article <DSAMPSON.90Jul16111525@x102a.harris-atd.com> 
dsampson@x102a.harris-atd.com (sampson david 58163) writes:
>In article <4995@milton.u.washington.edu>
>markad@blake.u.washington.edu (Mark Donnell) writes:
>
>>1) I am trying to print the output from WinWord on a remote NeXT
>>printer.  I print to a file....
>>I try printing it with 'lpr text.eps' - it spools, then dissapears....
>
>Mark -- The next time you print to a file with the windows postscript
>driver, take a look at the result.  The first line in the file will
>have something like this (without the quotes) "^D%! Adobe 2.0".
>The problem is that ^D as the 1st character.  

This much is correct.

>						The Apple printers want
>to see the %! as the 1st two characters in a postscript file...
>The %! tells the printer "Postscript commin' at ya".  The ^D in
>front of that signal confuses the printer.  It doesn't properly parse
>the line.  I believe that ^D is an attempt to force a top of page
>(i.e. kicks out a sheet).

This is a common misconception.  In fact, the ^D byte ($04) is a protocol
used by the serial and Centronics drivers in many PostScript printers, and
it means "end of job".  Other communications methods may have out-of-band
ways of conveying the same information: e.g. AppleTalk has an out-of-band
"end of job" signal and so does not intercept ^D bytes in the input stream.

The "%!" is a convention which means "this is a PostScript language program", 
rather than text, or PCL, or whatever.  It conforms nicely to the Un*x 
convention that the first two bytes of the file tell how to handle it -- 
remember "#!/bin/sh"?  Since "%!" starts with "%", it is a PostScript comment,
and the PostScript interpreter ignores it.  (There are other conventions; 
ask Adobe Developer Support for the Document Structuring Conventions spec.)

The standard Windows PostScript language driver, mixing protocol and page
description, emits a page description with ^D bytes before and after it.  If
you send this stream to a printer via a serial or parallel port, the
printer's RS232C or Centronics drivers interpret the initial ^D as the end
of a (null) job and remove it from the input stream.  The following bytes
are the next job.  The drivers interpret the final ^D byte as an end-of-job
signal, remove it from the input stream, and pass the end-of-job signal to
the interpreter.  The interpreter reverts to its initial state and waits for
the next job.

If you intercept the Windows driver's stream into a file, and print it with
the NeXT's 'lpr', it is sent to the Display PostScript interpreter via a
NeXT-internal method, with an out-of-band protocol for end-of-job signals.
(The same problem arises if your Un*x box sends the file to a printer via
AppleTalk.)  The ^D bytes have no special meaning, so they are passed to the
interpreter.  The interpreter treats the initial ^D byte as an undefined
PostScript name, and flushes the job with an error.

However, David came up with exactly the right workaround:

>I wrote a simple program in C to filter the Control-D out of the file.
>So my procedure for printing to a remote printer is as follows:
>
>1) Print to a file using the postscript driver from whatever windows
>application you're in
>
>2) Exit windows and filter out that ^D
>
>3) Issue the lpr filename command command to the remote printer.  

I hope this explanation is helpful.

>David Sampson                                         Harris Corporation
>dsampson@x102a.ess.harris.com                   Gov't Aerospace Systems Divison
>uunet!x102a!dsampson                                  Melbourne, Florida

	--Jim DeLaHunt		delahunt@adobe.com		    (Internet)
	  Technical Support	...!{sun,apple,decwrl}!adobe!delahunt   (UUCP)
	  Adobe Systems Inc.	+1-415-962-3790 (Phone)	 +1-415-961-3769 (Fax)