> From [EMAIL PROTECTED] Mon Oct 20 13:28:39 2003
> Date: Mon, 20 Oct 2003 22:28:30 +0200 (CEST)
> From: Henrik Edlund <[EMAIL PROTECTED]>
> To: Patrick Powell <[EMAIL PROTECTED]>
> Cc: [EMAIL PROTECTED]
> Subject: Re: LPRng: Brave souls needed to test SNMP Page Counting
>
> On Mon, 20 Oct 2003, Patrick Powell wrote:
>
> PP> SNMP runs over UDP. The network security issues can be hideous or
> PP> trivial, depending on the approach you take.
>
> Yeah. By using a private VLAN for the printers (like you have to do anyway
> to protect them from TCP :]), I believe this is not a security issue
> anymore. But it is something to consider as a con on a non-private VLAN of
> course.
>
> PP> If you use SNMP, then you have to fight through the various OIDS, and
> PP> then make sure that the values you get for 'end of job' really do
> PP> match the 'end of job'. You also have to make sure that the
> PP> Pagecounter value is really the right value.
>
> I guess you mean you have to be sure the job really has ended, that is
> wait for the printer to become idle (as if it is down it might have run
> out of paper and still want to continue printing later) and so on and so
> forth? And for the right value, you have to be sure it is counting
> impressions if that is what you want?
Correct. I have found that the mysterious 'other' status returned
by some printers (DeviceStatus) is really odd. It appears to be
the case when the printer is warming up (on some HP printers) and
when a job is being received on others.
>
> PP> If you use PJL/Postscript, then you have to fight through the various
> PP> PJL insanities, and hope that PJL Echo, PJL JOB/END, etc., work. If
> PP> you are using USTATUS, then you need to make sure that you really do
> PP> get status :-) You also have to make sure that the Pagecounter value
> PP> is really the right value.
>
> So the major thing with PJL is that the commands are not supported or do
> basically not work as specified? Going into PJL insanities, could you give
> more details on these? USTATUS being PostScript right? What else would you
> get if not status? Comparing these two options, is PJL more support/a
> better choice then PostScript, or would you say that you have to try with
> both options depending on printer?
The HP PJL Reference states categorically that 'PJL ECHO' is not
carried out until the previous job completes.
Well, it turns out that this is not true - if the previous job has been
rasterized but pages are still moving through the paper path, you get
an echo. Oh yes, and if you get a jam, all bets are off.
The PJL USTATUS command is supposed to return 'Unsolicited Status'
at a specified interval. It may or may not. If the printers is off line,
it sends an 'offline' ONCE on some printers. It may send an 'ONLINE'
or 'PRINTING' message when put back online.
>
> I have a pretty good idea of the success ratio of using SNMP and that you
> cannot trust that a printer supports (meaning returns intelligent values)
> even though the manufacturerar says so. (Canon being a bad one here and
> Xerox a good one.)
>
> What is the success ratio of PJL/PS? Beinh behind IFHP you should have
> some idea on the success people have had with it?
If you want to wait (i.e. - poll the printer periodically) you can get
very accurate page counts.
The PJL mechanism and the SNMP OID's are either both functional
or both broke.
If you need to get 'offline status' then SNMP is your only hope.
>
> PP> It is a case of death by fire or death by water.
> PP>
> PP> Patrick ("You get to choose") Powell
>
> Concerning your SNMP implementation in IFHP, you might want to read my
> experiences in my report later on (those I haven't shared so far) and look
> at my implementation, so you can work around them if you don't already
> have.
>
> There are some interesting special cases and solutions that I ran into
> when testing SNMP with 10 or so printers. I currently do not have time to
> dig into the SNMP test release if IFHP myself right now. :-(
>
> Henrik
>
-----------------------------------------------------------------------------
YOU MUST BE A LIST MEMBER IN ORDER TO POST TO THE LPRNG MAILING LIST
The address you post from MUST be your subscription address
If you need help, send email to [EMAIL PROTECTED] (or lprng-requests
or lprng-digest-requests) with the word 'help' in the body. For the impatient,
to subscribe to a list with name LIST, send mail to [EMAIL PROTECTED]
with: | example:
subscribe LIST <mailaddr> | subscribe lprng-digest [EMAIL PROTECTED]
unsubscribe LIST <mailaddr> | unsubscribe lprng [EMAIL PROTECTED]
If you have major problems, send email to [EMAIL PROTECTED] with the word
LPRNGLIST in the SUBJECT line.
-----------------------------------------------------------------------------
