http://www1.sqp.com/MasterIndex/installation_guide/installation_guide_005833b0.txt
has an install guide for the host sw.
msokolov_at_ivan.Harhan.ORG (Michael Sokolov) writes:
> PostScript interpreter. However, the firmware image for this VAX,
> containing the PS interpreter and everything else, does not reside in
> ROM but is instead down-line loaded on powerup over Ethernet, I presume
> via MOP.
The LPS20's, LPS32's and LPS40's I used to run indeed worked via MOP.
According to
http://h71000.www7.hp.com/wizard/wiz_5960.html,
an LPS17 can also boot via bootp.
> So here comes my first question to the list: does anyone have an LPS
> firmware image or is one publicly available on some archive site?
I probably have something (tk50 or cd) that has VMS LPS sw on it,
and I might have ultrix or osf/1-digitalunix-tru64,
but I'm not at all sure when I'd have time to dig it out.
It might well be on a dec consolidated sw distro if you could find
one; lots of people (perhaps including your local dec dealer...)
should have those, and be able to check the T.O.C. for you. In
addition to the OS image, there is a remote console program.
They seem to have sold the host sw for aix, hp-ux, sunos, solaris,
and some flavor of windows in addition to vms, ultrix, and osf/1-digitalunix-tru64.
> DEC docs seem to imply that the firmware is the same for all LPS'es, is
> this true or not? I need an image that can run on LPS17 as that's the
> only LPS I can get.
The images are different.
Only the LPS 17, 17/600, and 32 Plus could do Level 2 postscript.
Different machines had different cpu bus interfaces, ethernet
controllers, and numbers of input and output trays, and they were
based on at least three very different print engines. The 17/600 could
do 600dpi, while the 40 certainly could not, and the 17's also had HP
PCL5 ROMs in them, although I do not know if you could use the machine
as a pcl printer without going thru a network boot first.
> Second question. It is my understanding that once an LPS has booted its
> firmware, it becomes an independent node on the network accepting print
> jobs from anywhere on the network. Is this true or not?
yep.
though having a console and logging host isn't bad.
>... The protocols
> it speaks are DECnet and TCP/IP, right?
The 17 certainly does both, I'm not sure the 40 ever did IP.
>... If so, how does it obtain its
> DECnet and IP addresses? And if I want only one of the two, how do I
> configure it? Does it down-line load a configuration file from the MOP
> server along with the firmware image specifying DECnet and IP addresses?
> What is the format of this file?
Yep, it downloads a config file.
I'm not sure of the format, but the sw kit that contains
the firmware images and console program has a program that generates it.
> Third question. It is my understanding that the protocol spoken over TCP/IP
> is extremely simple: no LPR/LPD or anything like that, just a simple TCP
> port, you connect to it and everything sent to that TCP port goes to the
> PostScript interpreter's %stdin, and this port is 170. Is my understanding
> correct?
Varies by model, I'm pretty sure. See
http://h71000.www7.hp.com/wizard/wiz_3960.html
http://h71000.www7.hp.com/wizard/wiz_4045.html
http://h71000.www7.hp.com/wizard/wiz_3960.html
implication seems to be that the 17 does raw postscript on port 2501,
plus also speaking lpd at the usual place, and that port 170 is used
by a magic dec printer protocol that might or might not be raw postscript
with some wrapper and escape sequences added.
> Fourth question. Is this PostScript-over-Ethernet connection on TCP port
> 170 binary clean, or is 0x04 (^D) interpreted as end of one PS job and
> beginning of the next? In other words, does my lpd on the 4.3BSD host
> driving it need to send a 0x04 between print jobs or should it close the
> TCP connection and reopen it instead?
I would think that it would have to be binary clean, because otherwise
a control-d in some eps file would trash the connection; if one is speaking
postscript, I would expect a /showpage or somesuch to be end of job.
Of course, the implementation could be brain-dead, but I would hope not,
since they likely had postscript printer people making these things,
and not the rather iffy ultrix folks.
> Fifth question. Is there any access control mechanism by which LPS firmware
> can be configured to accept connections only from certain sources, or will
> it always accept connections (and print jobs) from the entire Universe?
Seems to be allow and deny lists for decnet and ip.
--akb
Received on Wed Jun 16 2004 - 03:38:02 BST