Path: utzoo!mnetor!uunet!husc6!mailrus!umix!umich!mibte!gamma!sabre!sword!faline!bellcore!buzz!nrh
From: nrh@buzz.bellcore.com (Nat Howard)
Newsgroups: comp.lang.postscript
Subject: PostScript(tm)  Validation Efforts?
Message-ID: <6322@bellcore.bellcore.com>
Date: 24 Mar 88 17:03:07 GMT
Sender: news@bellcore.bellcore.com
Reply-To: nrh@buzz.bellcore.com (Nat Howard)
Organization: Bell Communications Research, Morristown NJ
Lines: 26

Now that there are several PostScript clones around or about
to be around, we printer-buyers have a problem:  How do you know
if a clone measures up?

When I first thought of this, the only answer I could see was to 
try printing stuff on different printers to see what would happen, but this
has some difficulties: visual fatigue, human error, blah, blah...

It then occurred to me that one might be able to do better: the
"framedevice" operator is given a proc as argument that is called
""to cause the contents of the frame buffer to be transmitted to the 
physical output device" (PLRM pp. 162)

If we knew what this proc was expecting as arguments, we could maybe
arrange for it to send a representation of the raster frame buffer
back down the communications link, and then do something (heaven only
knows what) to verify that you get the same bits out when you
put the same bits in.  

Has anyone tried this approach, or come up with something else
in the way of validation?  I agree that the resultant bitmaps
(even at 300 DPI) need not *exactly* match between an Adobe Postscript
and a clone printer -- particularly if they use different
technology engines (so that one would want some difference).  On the
other hand, the relative darkness of different areas of the pages
should be alike, and one could go on from there.