A question was posted recently to comp.unix.osf.osf1, and even though it wasn't
sent (as far as I'm aware) to the alpha-osf-managers list, I have seen similar
sorts of questions before, so I decided it might be worthwhile to send my reply
to this list, in addition to the comp.unix.osf.osf1 gang and to the person who
originally posted the question. So here goes ...
>> We recently added an HP 1600CM printer with a JetDirect card, and are
>> having some problems with lpr. Jobs sent to the printer will print OK,
>> but the queueing is not working properly. The symptoms are that 1) When a
>> job is printing, lpq just hangs and 2) Sending more than 1 job at a time
>> results in an error page being printed and losing both jobs.
[ Beware ... long and intermittently flamish reply follows! ]
We've had a LaserJet 5 Si/MX for nearly a year, with similar problems with its
JetDirect card. The executive summary is ...
mechanically, the printer is superb
but
the firmware is an abomination.
Let me elaborate. (There may be hope for a (partial) solution to your problem.
Read through to the end!)
The symptoms we encountered, almost from the start, included:
-- If multiple jobs were queued to the printer, the first would print OK, but
the jobs queued behind it would vanish. (Similar to the problem you described,
but not quite as bad.)
-- An "lpq" command would work fine if the queue were empty, but would hang if
a job were being printed. (Sounds identical to your problem.)
-- Various examples of PostScript streams crashing the printer (turning on the
red light, displaying an error code in the panel display, which, when looked up
in the users guide, would imply that this was a "never should happen" situation
and that HP should be contacted).
To make a VERY long story short, what I discovered is that the first problem,
as well as most instances of the third, went away if I enabled printing of a
banner page. The printer had come with this enabled by default, but I turned
it off immediately, to keep from needlessly wasting paper. After many months
of frustration (and a couple postings of my own), I got a message from someone
(I've forgotten who, I'm afraid, so I can't give credit appropriately), which
suggested that enabling the banner page "solved" a lot of problems. I tried
it, and things are much better now.
The "lpq hangs if the printer is busy" problem is still present, but at least
we can queue multiple jobs without having #2 through #N disappear. However, we
are wasting a lot of paper needlessly.
I suspect that a lot of the firmware bugs might be a side-effect of the fact
that HP may be doing a lot of its debugging with banner pages enabled, since
in that mode they wouldn't have encountered a lot of the bugs that we've seen.
One other long-standing gripe is that the service people whom HP has sent out
when we've reported these problems are almost without doubt the most clueless
ones I've ever met. (And hey, we've seen some pretty dreadful ones from the
various computer vendors!)
For example, one of the times when I had a PostScript stream which reliably,
and reproducibly, crashed the printer, I started describing the problem to
the guy when he arrived. After the initial "are you using a PC or a Mac?"
question that seems inevitable in these sorts of situations, followed (after
explaining that it wasn't either, it was a UNIX workstation) by several more
equally off-target questions ("what application are you running on your PC?")
it became clear that this guy ...
-- had barely heard of UNIX;
-- hadn't the foggiest idea what lpd emulation was;
-- and (brace yourself) had no idea what PostScript was.
Unfortunately, HP's organizational structure seems to have no facility for a
customer to submit a bug report, other than telling a service person who is
totally incapable of understanding the problem. After a number of telephone
calls, I did finally find someone in HP tech support who, though not really
his job to do so, was willing to serve as a conduit for information (in both
directions), but I'm reluctant to pester him too frequently, because he has
other things he's supposed to be doing (and no, I'm *not* going to give out
his e-mail address!).
I admit that some of the problems are subtle (for example, one of the bugs
we stumbled across would cause a printer crash only when three conditions
were true simultaneously: (1) duplex mode enabled; (2) page count is odd;
(3) PostScript stream has a "paper selection by tray" or "paper selection by
media type" command). Others, however (such as the one where jobs vanished
if queued while a printer was busy, and the banner pages were disabled), or
the "lpq hangs when printer is busy") are simply inexcusable.
One other approach I've considered is using FreePort Express to try porting
HP's JetAdmin software to DUnix. HP makes the software available on their
web site, but there are versions only for HP/UX and SunOS-Solaris. It may
be worth noting that one of the README files says that ...
The JetAdmin Utilities for UNIX product provides a
robust printer management solution for Hewlett-Packard
printers network connected with HP JetDirect print
servers.
Is this a subtle acknowledgment that their lpd emulation does *not* provide
a robust solution? Hmmm ...
(The JetAdmin stuff communicates with the printer on a totally different TCP
port (9100, I think) from what lpd uses.)
By the way, the man pages which come with the JetAdmin package drop a hint
about why "lpq" hangs when the printer is active. It says that an
HP network printer grants one TCP connection at a time
Could this imply that when a local lpd is talking to the printer's lpd code,
the printer won't accept a TCP connection from an "lpq" command? Sounds like
it, doesn't it. Bleah.
Anyway, if you have banner pages disabled, try enabling the feature. It may
make some (but definitely not all) your problems disappear. And, if you've
already got banner pages enabled, then I have no useful suggestions, unless
you'd like to try doing an FPX port of the JetAdmin package yourself. We'll
all be interested to hear how well it works (or doesn't work). If and when
I ever get around to doing it, I'll post a followup; but I doubt I'll have a
lot of time to devote to it in the immediate future.
Mark Bartelt 416/978-5619
Canadian Institute for mark_at_cita.utoronto.ca
Theoretical Astrophysics
http://www.cita.utoronto.ca/~mark
"Sheep not busy being shorn are busy frying" -- Dylan, at a NZ lamb barbecue
[ singing "It's all right, ma (I'm only bleating)" ]
Received on Wed Mar 19 1997 - 00:17:52 NZST