🚀 go-pugleaf

At my office we have a new LaserJet III with PostScript cartridge, and a
286-clone running Windows 3.0 and some Windows applications (PageMaker,
Corel Draw, PowerPoint). The computer (and a couple of others) are run
through an automatic switching box (CCAT-4, no buffer, using serial
cables from the computers).
Some of you have already guessed the problem: we can't print PostScript
files through Windows 3.0 and the switching box. We get a printer error
indicating that a non-PostScript character has been sent, and the
printer stops. We can print PostScript documents fine through a
non-Windows application (WordPerfect 5.0), and we can print everything
if we connect the computer directly to the printer, bypassing the
switching box.
I have had several conversations with both Microsoft and the
manufacturer of the switching box, and found a batch of letters to the
editor in one of the desktop magazines from frustrated users having
similar problems. The problem apparently is not restricted to LaserJet
III's, but common to any PostScript printer hooked up to a switching
box. Basically, Windows 3.0 doesn't like switching boxes,
and the Microsoft people had few useful suggestions about what to do
about it. Their whole attitude is that its not their problem, is the
switching-box people's problem, and they say there are no plans to fix
this in future versions of Windows.
Now, it appears that this is not true of ALL switching boxes. One theory
floating about is that parallel switching boxes with buffers can handle
it. The Microsoft people said that may be true for some such devices,
but probably not all, and they had no info on which brands might be
better. The fact that we can print to a file, then go to the DOS prompt
and copy the file to LPT1 (which is in turn rerouted to COM1 and hence
to the switching box and then the printer), and get our PostScript
output, tends to suggest that the serial interface is part of the
problem for the WIndows/switching box interaction. Though as I said, we
can print directly from Windows to the printer, through
LPT1-routed-to-COM1-and-a-serial-cable, just fine.

SOOO....has anybody else out there run into this problem? Found a
solution? Does anybody know of a switching device (buffered or not)
which has been proven to be able to handle this?

prb@chinet.chi.il.us (Paul Botts)