[ILUG] remote print failure

[Peter Flynn]

     on a client machine its setup as a postscript printer so it sends
     postscript.  The local machine has it as a LaserJet4, so sends as
     HP format ?  Heres an entry from me printcap.

oops. My fault, I should have seen that. I've set it back to a
PostScript printer but it still says "job could not be printed".
printcap now says

##PRINTTOOL3## LOCAL 
lp:\
	:sd=/var/spool/lpd/lp:\
	:mx#0:\
	:sh:\
	:lp=/dev/lp1:

I tried adding rp=raw but it had no effect. 

   firstdup:\
	   :sd=/var/spool/lpd/firstdup:\
	   :mx#0:\
	   :lp=/var/spool/lpd/firstdup/null:\
	   :sh:\
	   :lf=/var/log/lpd-errs:\
	   :rm=firstdup.hdl.ie:\
	   :rp=raw:

      I've got no filters setup on this, so the lpd sends stuff as it
      recieves the stuff, no matter what format the stuff is in.

Nope, I've tried it like that, with no filter, and it still says
"job could not be printed". It says that for all imaginable settings.
Why on earth doesn't it emit a sensible error message?

This is crazy. I've had that printer running on a remote Alpha under
Digital Unix 4 for a couple of years with no problems, printing from
the client I'm using (so on there all I did was change the machine
name to the new server name:

castor:\
        :rp=lp:\
        :rm=castor:\
        :sd=/usr/spool/lpd:

The problem seems to be the Linux server's reluctance to print the
file as instructed, or to give a meaningful error message. Just
saying "could not be printed" is unutterably ridiculous. 

      Note if you have more than one printer have different spool
      directories.

Nope, only one printer.

      Heres an entry from a client machine for the above printer:
      ##PRINTTOOL3## REMOTE POSTSCRIPT 600x600 a4 {} PostScript Default 1
      lp:\
	   :sd=/var/spool/lpd/lp:\
	   :mx#0:\
	   :sh:\
	   :rm=portia.hdl.ie:\
	   :rp=firstdup:\
	   :if=/var/spool/lpd/lp/filter:

Why should I want to filter it at the client end if I'm sending PS
from the client to a server set to print PS?  I don't have access to
any filters that I'm aware of on the client anyway (a Sun running
SunOS 4.1.3) and like I said, it's been sending to that printer
attached to an Alpha faultlessly for ages...

   > I don't think hosts.allow is relevant for lpd.

      you're right, I'm getting mixed up.   hosts.equiv then ?

Same as hosts.lpd -- that's working, cos it's receiving the job. It's
just refusing to print it.

///Peter