[comp.sys.dec] LJIIID w/HP PS cart. on a DECstation 2100

steven@uicadd.csl.uiuc.edu (Steven Parkes) (02/07/91)

We're running our LJIIID using a true-blue HP PS cartridge on a DECstation 2100
using transcript.  There seems to be some difficulty with this setup:

1) 19200 BAUD : doesn't work at all.  Printer immediately comes up with an
   I/O Config Error on the front panel.  The transcript package logs the
   error from printer, but its completely garbage.

2) 9600 BAUD : prints, but comes up with the I/O CONFIG ERROR periodically.
   When this happens, the printer log file gets

		%%[ PrinterError: Miscellaneous Error ]%%

   Pushing `continue' helps some (it goes on) but the error reoccurs.
   Eventually, some sort of PS ERROR ## occurs and the log file gets

	%%[ PrinterError: Miscellaneous Error ]%%
	%%[ Error: undefined; OffendingCommand: 439m ]%%
	%%[ Flushing: rest of job (to end-of-file) will be ignored ]%%

   The position of this seems pretty much random -- the same PS can be sent
   again and may work.

3) 4800 BAUD : pretty much the same as 2), though possibly happens less often.

This seems to be unaffected by settings of PAGE PROTECT, AUTO CONF, or
ROBUST XON.  I've been running in DUPLEX mode -- I don't know if single
sided affects this.  The printer has a total of 5M.

Wonderfull printer, but this is very annoying.  I haven't been able to
figure out from the printer what the problem is.  I'm not sure if its
flow-control related.  We have another non-LJ PS printer attached to another
2100 and it has no problems running at 19200.

Any suggestions would be welcomed.

steven parkes ---------------------------------------
University of Illinois Coordinated Science Laboratory
steven@pacific.csl.uiuc.edu -------------------------

steven@uicadd.csl.uiuc.edu (Steven Parkes) (02/09/91)

|> 1) 19200 BAUD : doesn't work at all.

I've since been told that 19.2 doesn't work on DECstations and went back and
found that our other printer is indeed running at 9600.

The concensus seems to be that the DECstation doesn't respond to XOFF
quickly enough and thus overflows the LJ's buffer.  This is going to be
damn hard to fix, I'm afraid.

In looking through the tty driver documentation, the ULTRIX version talks about
an AUTOFLOW mode in the local mode word.  Supposedly, it does flow control
in hardware and therefore could be expected to respond to XOFF immediately.  I
tried modifying  the printcap to:

	:br#9600:rw:fc#0000336:fs#0000041:xc#0:xs#0041040:

but it didn't seem to help.  Does anyone know if the hardware suport for this
exists in DECstations or is only on VAXes, etc.

Any other suggestions would be greatly appreciated.

steven

steven@uicadd.csl.uiuc.edu (Steven Parkes) (02/13/91)

I have received a patch from the DEC Support Center that speeds the response
of the DECstation to the XOFF generated by the LJ.

The following patch is to be applied to a 4.0 kernel.  I don't know if the
same patch is needed for 4.1.  I've applied it to my system and it seems to
eliminate the problem.

steven parkes ---------------------------------------
University of Illinois Coordinated Science Laboratory
steven@pacific.csl.uiuc.edu -------------------------

*** /usr/sys/machine/common/conf.c.orig	Tue Feb 12 10:46:23 1991
--- /usr/sys/machine/common/conf.c	Tue Feb 12 10:47:53 1991
***************
*** 1072,1077 ****
--- 1072,1078 ----
  
  #ifdef mips
  int	cnopen(),cnclose(),cnread(),cnwrite(),cnioctl(),cnselect();
+ int	cnstop();
  #endif mips
  
  #ifdef vax
***************
*** 1896,1902 ****
  #endif vax
  #ifdef mips
  	/* d_ttys will be filled in by cninit */
! 	cnioctl,	nulldev,	nulldev,	0, 
  	cnselect,	nodev,		0,	0},
  #endif mips
  	{dzopen, 	dzclose,	dzread, 	dzwrite,	/*1*/
--- 1897,1903 ----
  #endif vax
  #ifdef mips
  	/* d_ttys will be filled in by cninit */
! 	cnioctl,	cnstop,		nulldev,	0, 
  	cnselect,	nodev,		0,	0},
  #endif mips
  	{dzopen, 	dzclose,	dzread, 	dzwrite,	/*1*/

wnp@iiasa.ac.at (Wolf PAUL ) (02/14/91)

In article <1991Feb6.192634.29403@roundup.crhc.uiuc.edu> steven@pacific.csl.uiuc.edu writes:
>We're running our LJIIID using a true-blue HP PS cartridge on a DECstation 2100
>using transcript.  There seems to be some difficulty with this setup:
>
>1) 19200 BAUD : doesn't work at all.  Printer immediately comes up with an
>   I/O Config Error on the front panel.  The transcript package logs the
>   error from printer, but its completely garbage.
>
> ... (rest of error description deleted)


We have not yet tried using the HP-PS cartridge with our LJIIIs, but
had the same problem trying to use the PacificPage Postscript
cartridge with a LJII: works fine when the printer uses the parallel
port, and not at all when using the serial port.

Maybe a serial-to-parallel converter is the only solution?
--
W.N.Paul, Int. Institute f. Applied Systems Analysis, A-2361 Laxenburg--Austria
PHONE: +43-2236-71521-465            INTERNET: wnp%iiasa@relay.eu.net
FAX:   +43-2236-71313                UUCP:     uunet!iiasa!wnp
HOME:  +43-2236-618514               BITNET:   tuvie!iiasa!wnp@awiuni01.BITNET