Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!usc!zaphod.mps.ohio-state.edu!sol.ctr.columbia.edu!emory!hubcap!ncrcae!ncrlnk!rd1632!hann
From: hann@rd1632.Dayton.NCR.COM (Ed Hann)
Newsgroups: comp.sys.ncr
Subject: PostScript Printer gettydef
Message-ID: <1014@rd1632.Dayton.NCR.COM>
Date: 7 Mar 90 19:42:34 GMT
Reply-To: hann@rd1632.Dayton.NCR.com (Ed Hann)
Organization: NCR Research & Development, Dayton, Ohio
Lines: 92

I just finished configuring a Tower to use a PostScript printer on a serial
port, and discovered a strange situation.

The Tower is a Tower 32/600 with Tower OS 3.00.01.

The PostScript printer is an NCR 6416 laser printer with the contoller
replaced with a PSJet+ controller.  This makes it equivalent to a QMS 800+
laser printer.  The printer is configured to use the 25 pin channel: 9600
baud, parity ignored, DTR protocol.

I used a cable wired like a 1308-C037/C036 Intelligent Terminal Cable, as
described in NCR TOWER Series Site Preparation Guide, page B-13, which says
this will work with an NCR 6416 Laser Printer.

I experimented with different parameters for the gettydef entry in
/etc/gettydefs that is combined with the a model /usr/spool/lp/model/laser to
create the printer interface file /usr/spool/lp/interface/printer2.  I
finally got the following getty definition to work:

	laser# B9600 CLOCAL CS8 ICANNON # b9600 CLOCAL CS8 ICANNON # # laser
                                   ^^
Please note the misspelling        ||.  I inadvertantly mistyped the
parameter.  The documentation spells it ICANON.  Other entries in gettyfefs
also spell it ICANON; yet it works.

I didn't discover this until I was configuring an identical printer with an
identically wired cable on a Tower 32/600 with Tower OS 1.03.02.  I used the
proper spelling in /usr/spool/lp/local/lpgettydefs/laser (the gettydef file
is located in a different place in 1.03.02).  I used the same model as that
used on the previous Tower.

Everything appeared the same, but the printer would only flash briefly, and
no print would come out.  I concluded that since I had misspelled ICANON on the
other Tower, the parameter was not needed.  Without the parameter I got the
same symptoms, so I added it back, but used the incorrect spelling ICANNON;
and it now works.

WHY?  Is the documentation wrong?  Is ICANNON actually right?  Does something
good happen with misspelled parameters?

The following code is a copy of the model I used on both Towers:

########################### Model begins here ###############################
# lp interface for psjet+ laser printer
#
copies=$4
shift; shift; shift; shift; shift
files="$*"
i=1
while [ $i -le $copies ]
do
	for file in $files
	do
		echo "\004\c"
		cat "$file" 2>>/tmp/laser$$
		echo "\004\c"
	done
	i=`expr $i + 1`
done
mail lp </tmp/laser$$
rm -f /tmp/laser$$
exit 0
########################### Model ends here ###############################

The following represents the wiring diagram for the cables:

		DB-15 Tower			DB-25 Printer

                                           |---  (Connect to Shield)
                                           |
	                                   |---   1 Frame Ground

	Transmit Data	 1 --------------------   3 Receive Data

	Receive Data     9 ---------------------  2 Transmit Data

	Request to Send  2 ---------------------  5 Clear to Send

	Clear to Send   10 --------------------- 20 Data Term. Ready

	Data Term. Ready 4 -----------------|---  6 Data Set Ready
                                            |
	Data Set Ready   3 ---|             |---  8 Carrier Detect
                              |
	Carrier Detect  12 ---|-----------------  4 Request To Send

	Signal Ground   11 ---------------------  7 Signal Ground
-- 
Ed Hann					(Ed.Hann@Dayton.NCR.COM)
Applied Research, R&D Division		(1 513 445-1061)
NCR Corporation
1700 Patterson Blvd., Dayton Ohio 45479