Path: utzoo!attcan!uunet!tut.cis.ohio-state.edu!ucbvax!agate!violet.berkeley.edu!izumi
From: izumi@violet.berkeley.edu (Izumi Ohzawa)
Newsgroups: comp.lang.postscript
Subject: Re: Real bottleneck 1200 bps on LaserWriter NTII
Message-ID: <1990Jan13.195339.16434@agate.berkeley.edu>
Date: 13 Jan 90 19:53:39 GMT
References: <1731.25ae524c@cc.helsinki.fi>
Sender: usenet@agate.berkeley.edu (USENET Administrator;;;;ZU44)
Organization: University of California, Berkeley
Lines: 33

In article <1731.25ae524c@cc.helsinki.fi> tarkkonen@cc.helsinki.fi writes:
>1200 BAUD A BOTLENECK IN LASERWRITER NTII
>
>The department has one Apple LaserWriter NTII connected to a Unisys PW300
>through the serial interface.  The transfer of the data is quite unreliable,
>and using the speed 9600 baud will cause io errors about every second page.
>If the speed is redused to 1200 bauds the operation is faultless.
>
>The programs we use write to the LPT1 port and it is redirected the normal
>way  using: MODE LPT1:=COM2:.
>
Looks like a problem with improper handshake.  Out of the box, PC
(clones) and MSDOS cannot handle software handshake (XON/XOFF or
Control-S/Control-Q) to throttle transimission to the serial device.
On the other hand, out of the box, LaserWriters are configured to
do this software handshake.

Solution:
Look in the LaserWriter (NTII) manual for the section describing
serial port handshake.  There should be a short command sequence
for changing the default software handshake to hardware handshake.
Choose the command to set it to DTR/... handshake.

Since this command sequence is so short, you should be able to
send this sequence by doing a simple copy:
	copy commandfilename LPT1
or	copy commandfilename COM2 (if not redirected).

Make sure also that the cable has a wire going from Pin-20 (?)
on the LaserWriter side (DTR wire).

Izumi Ohzawa
izumi@violet.berkeley.edu