I apparently don't know how to select the parallel port after the bytes
get to the Netgear server.
Has anyone tried this, and if so, could I take a look at your "/etc/printcap"
file ?
Thanks!
73 de Bill W5GFE
b...@cs.ecok.edu
System = SCO_SV
Node = froghill
Release = 3.2v5.0.6
KernelID = 2000-07-27
Machine = i80386
BusType = ISA
Serial = 1NC057788
Users = 2-user
OEM# = 0
Origin# = 1
NumCPU = 1
>I'm trying to run a NetGear PS 110 Print server. It is a 2-parallel
>port printer server. I can't make either port work through "printcap"
>on my SCO box. I can print to the printers via the "setup" menu on a
>PC which runs the Netgear configuration software, and I can "ping" the
>IP which I have set on the Netgear box. When I print via the spooler,
>there is network activity, but the bytes don't make it out the parallel port.
>
>I apparently don't know how to select the parallel port after the bytes
>get to the Netgear server.
>
>Has anyone tried this, and if so, could I take a look at your "/etc/printcap"
>file ?
>
>Thanks!
>
>73 de Bill W5GFE
>b...@cs.ecok.edu
When you give up on lpr/lpd, try "Netcat". It works.
http://www.cruzio.com/~jeffl/sco/lp/netgear.htm
http://pcunix.com/KevinSmith/netcat/
http://www.tkrh.demon.co.uk/netcat.html
http://pcunix.com/Jeffl/printports.html
I have a PS104 and have installed several PS110 print servers at
customers.
--
Jeff Liebermann 150 Felker St #D Santa Cruz CA 95060
(831)421-6491 pgr (831)426-1240 fax (831)336-2558 home
http://www.LearnByDestroying.com WB6SSY
je...@comix.santa-cruz.ca.us je...@cruzio.com
I gave up on lpr/lpd before I heard from you, but thanks for the reply.
I did, indeed, use "netcat" -- which provided a great solution!
I finally wrote a new "model", with hints from several other sources including
Tony's pages, and used it to make things work. The scoadmin utility
was generally a wash-out, but installing the thing with lpadmin was
a cinch.
I added the IP address for the netgear PS110 to /etc/hosts
used scoadmin printer to add a remote unix printer with the netgear as
the host (remember to deselect the option which gives the extented
remote printing protocol!!!)
For simple setups this is all you have to do!
If you have more than one printer on your PS110 you will need to edit
/etc/printcap and change the
rp=
to P1 or P2 to indicate which of the two physical parallel ports on
the netgear you are refering to.
You can also use your Windaz Firstgear Software to configure logical
ports on the Netgear PS110 with printer initialisation strings. If
you do this then on the sco box you need to edit /etc/printcap and
change rp= to L1, L2,L3......L8 depending on which logical printer
port you want to refer to.
James Szabadics
True. But.. netcat works better.
http://pcunix.com/SCOFAQ/scotec7.html#getnetcat
You should maybe edit the sample printer interface script taking into
account the following things:
1 - "ffstrip" is required by the script but not found on my system.
2 - when you use the -oraw option, it's not really raw, it still sends
some init codes.
attached is my take on the above.
I also made a copy of sco's HPLaserJet which preserves everything from the
original and only adds the netcat stuff. In other words, if you had a
printer set up and working using HPLaserJet, it should work the same using
HPLaserJet.nc without any pain in the neck surprises. (as long as you
remember that any LPD printers were not using their interface script
before, and now they will.)
I have doubts about the legality of posting that script, but the changes
were simple enough I think I can make a script that will create a
HPLaserJet.nc by reading (and injecting stuff) your own HPLaserJet. Too
bad diff or patch are not part of the base system...
--
Brian K. White -- br...@aljex.com -- http://www.aljex.com/bkw/
+++++[>+++[>+++++>+++++++<<-]<-]>>+.>.+++++.+++++++.-.[>+<---]>++.
filePro BBx Linux SCO Prosper/FACTS AutoCAD #callahans Satriani
Thanks to diff -e ...
attached is a script to create a HPLaserJet.nc from HPLaserJet on a OSR
5.0.6 box.
it uses "ed" and since this was done on a OSR 5.0.6 box, I can't say if it
will produce a good result on older versions.
I don't think HPLaserJet has changed in a long time, so it might work on
5.0.5 and 5.0.4, I haven't checked. In any event, it won't hurt if you run
it on 5.0.4, it will just create a new file called
/usr/spool/lp/model/HPLaserJet.nc and that file may or may not be junk,
but it won't alter your existing HPLaserJet. I'd make a copy of it first
anyways though.
just run it by doing "sh mkhpljnc.sh"
it uses explicit paths so it doesn't matter where you are when you run it,
you should be root though.
The new interface script HPLaserJet.nc uses the same /etc/printers.nc that
the other netcat script uses (ahh that's another difference, sorry, I
prefered to use /etc/printers.nc instead of /etc/printers for this.)
Tony Lawrence wrote:
> True. But.. netcat works better.
> http://pcunix.com/SCOFAQ/scotec7.html#getnetcat
I have used LPR on many systems, without problems. Why is netcat better?
I have seen many posts advocating it use, but no simple explanation of why.
If you haven't experienced problems, you just haven't done enough :-)
With lpr or the hpjet stuff you'll get mysterious hangs, jobs not
printing. I suspect I have usually seen this at larger installatioons
with many printers, but I know I have also seen it with as few as a dozen.
I don't know why these things get whacked, and (since netcat) I don't
care. I use it automatically now, but of course every now and then I
get called into someplace new and they complain of printing problems.
Netcat is always on a cd in my pocket, and it always cures the problem.
--
Tony Lawrence
Methinks you have to roll back the clock about 4-5 years, when OSR5
LPR/LPD was seriously buggy. The most common problems that I saw were
hangs when printing a large number of small jobs, and high cpu
utilization. I've experienced both with LPR/LPD.
The alternative, HPNP for HP print servers, was worse. It had locking
problems, would leave garbage all over /tmp, would hang if you print
to an off-line printer, and belch bogus error messages if SNMP status
were not available from the print server. I also had some difficult
to reproduce problems that resulted in continuous retries with a
corresponding deluge in network traffic.
In 3.2v5.0.5, there were also some "scoadmin" printer configuration
problems that made printers disappear if added after an LPR/LPD
configured printer. lpstat -t would sometimes fail to show all
printers and often stall for extended periods waiting for a
configured, but powered off LPR/LPD printer, to come online.
AFAIK, the LPR/LPD, scoadmin, and lpstat problems were solved in some
update to 3.2v5.0.5 and remain fixed in 3.2v5.0.6.
However, 4-5 years ago, there was no suitable solution to the reliable
printing problem. I was looking for something that preserved the
print spooler system and the print spooler scripts, but communicated
directly with the print server. Kevin Smith supplied the answer with
his "netcat" program. I'm not much of a programmist, so additional
help was needed to modify the scripts to use netcat and to add the
/etc/printers configuration file.
I have a customer that was printing about 2000 checks per day. The
program creates on print job per check resulting in a huge number of
print jobs that had to work perfectly. We started with LPR/LPD, but
that kept hanging. We switched to HPNP, but ran into a wide
assortment of stupid problems (inability to recover from an off-line
condition, print server bugs, trash in /tmp, job counter errors, etc).
The system would also bog down somewhat when printing. When I setup
netcat, the hangs stopped, off-line recovery functioned, the print
server bugs were not an issue, and nobody even notices when it was
printing. We also noticed an unexpected printing speed improvement as
there is apparently some overhead to setting up a print job using
LPR/LPD and HPNP in the print server.
Try it. You'll probably like it.
I'm having problems like that today on a 5.0.6 that is less than a year
old and has 5.0.6a an any other patches that looked remotely applicable
applied. LPD did work ok for a while, then it started simply not printing,
as if it didn't think the printer or the print server was on line.
printers were physically ok, on-line, not jammed, people could print from
windows to them, power-cycling the printers and the print servers didn't
help, stop-start lpshut/lpsched didn't help... netcat worked immediately.
I should never have bothered with lpd...
Tony Lawrence wrote:
> Robert Steinmetz wrote:
>
>> I have used LPR on many systems, without problems. Why is netcat better?
>> I have seen many posts advocating it use, but no simple explanation of
>> why.
>
>
> If you haven't experienced problems, you just haven't done enough :-)
Possibly most of my installations as small with no mare that a few,
relatively lightly used printer.
>
> With lpr or the hpjet stuff you'll get mysterious hangs, jobs not
> printing.
I don't use the hpnp stuff because I have experienced those problems.
> I suspect I have usually seen this at larger installatioons
> with many printers, but I know I have also seen it with as few as a dozen.
>
> I don't know why these things get whacked, and (since netcat) I don't
> care. I use it automatically now, but of course every now and then I
> get called into someplace new and they complain of printing problems.
> Netcat is always on a cd in my pocket, and it always cures the problem.
>
Thanks to you and others who responded. I'll try netcat real soon now.
> --
> Tony Lawrence
>
I just realised it does *not* reproduce the script I wrote.
apparently some of the lines in my script cannot be automatically
reproduced using the "diff -e" --> "ed" mechanism.
when I was testing it, I didn't look down through the whole file when
looking at the generated result. my bad.
I bet the problem is the fact that I made the ed script into a "here"
document, and as such the shell is doing expansions & substitutions, so
probably I just need to back-slash a few dollar signs and back-ticks.
watch this space... (or don't wait for me and just follow this hint and
maybe fix your own copy if my guess is right)