HP Laserjet problem - solution & summary

[Hurf Sheldon]

 Many thanks for the several replies. We had not tried litout as the 
documentation said only the 'new' driver read the local bits and an
 'stty everything >/dev/lj' showed "old" after lpr ran so...for
whatever reason setting LLITOUT does work & was the crux of the problem.

 Chris Torek provided the secret switch of twiddling the baud rate to
get the local bits read as we were doing manual changes to see what would do
what. The final solution was not to use a filter but to use the following:

  We installed Ron Flax's 'Laser' program to use as a job handler
(very kindly sent in response to our request by several readers)
We included '\033&k2G' in the printer setup line to map cr->cr-lf,ff->cr-ff
 - all works well ( we use the HP for graphics & text, so far not for
 nroff or ditroff) 

The lpr command in 'laser' is set to 'lpr -Plj -h' as the burst page
gets prepended & not run through 'laser'. For graphics output we prepend
the printer setup to the file & send it through lpr with a script:
      !# /bin/sh
	echo -n "^[E" >/tmp/prep
	echo -n "^[&k2G" >>/tmp/prep
	cat /tmp/prep $1 | lpr -Plj -h
	rm -f /tmp/prep
 
A variation of this is installed as 'plj' in /usr/local for sending completely
unfiltered jobs to the HP (ie: files written with HP codes already imbedded)	
important printcap settings:
			    fc#0177777 (clears register)
			    fs#0000321 (no parity, crmod, tandem)
			    xc#0       (clears register)
			    xs#0040040 (LDECCTQ,LLITOUT)
see tty(4), printcap(5) for explanations

 BTW: our device is /dev/lj, owner daemon, created with 'mknod', not 'ln'
   (as per software notes in Ultrix1.2 distribution)
 I will repost the Laser source if requested. 

 As a notable aside - we were switching cables between computers & the printer
to see if modem control would solve our problems & the printer refused
to communicate any more. This made us crazy until we found that the line 
receiver on the printer had died. The local HP dealer said they would only
do board swaps & would have to order a board. Time & money! We took a deep 
breath & took the printer apart - one 60cent chip later we were back up.
My guess is that static build up in the cables or whatever killed the chip
so learn from our mistakes & turn what you can off & static ground the rest
when swapping cables.  

 Many, many thanks


     Hurf Sheldon			Arpa.css: hurf at ionvax.tn.cornell.edu
     Lab of Plasma Studies	
     369 Upson Hall			phone: 607 255 7267
     Cornell University
     Ithaca, N.Y. 14853