P1394 Mail Archive: Re: P1394> units and logical units definition

P1394 Mail Archive: Re: P1394> units and logical units definition

Re: P1394> units and logical units definition

Atsushi_Nakamura (atsnaka@bsd.canon.co.jp)
Tue, 17 Mar 1998 10:30:07 +0900

Eric wrote ;
> >
> > The other model is difficult if you want to have multiple
> > independent units in a node, because there is no clear way
> > to match up which Device Status Unit Directory goes with which
> > Device Control Unit Directory, or Print Job Unit Directory.
> > Rev 0.1d of the profile doesn't show how to do this (or I
> > missed it...)
Greg wrote ;
> No, you didn't miss it. It is not there. This may be an issue.

It is SORT of there....

This can be handled with FDS, the Function Discovery Service
discussed in p1212r, IF we want to match units up by (physycal)
function instances.
(although it is not the original intent of FDS..)
FDS is a ROM structure that acutally matches up an functional
physical instance with all of it's associated (unit) directories.
The problem is that there is no telling whether those multiple units
have to be "activated" simultaneously in order to achieve a function,
or just "choices"(ex. "A function can either talk in SBP-2 or DPP").

Rev0.1d states the use of FDS, so it's SORT of there...

Eric worte ;

> Even if each of those models is allowed by SBP-2, I think
> the first one (a single unit with multiple LUNs) is both more
> in the spirit of SBP-2 and more in the spirit of 1394.

I do not know if it is right or not, but this is a similar
response I get from 1394-ers around me.
I have a hard time explaining that a 1394 printer (will probably)
consist of 2 or more unit (directories).
Most people had imagined a printer with only 1 unit with 1 LUN.

Because there still is some talk about desires for mutiple
independent channels within 1 LUN even inside PWG, I think this
matter is worth going back to, get in synch and confirming again.
(We forgot to give examples the last time, I think Greg's examples in
the last few mails helped some people to get the idea, Thanks Greg.)

On radical comment I got was,
If the printer was going to consist of multiple LUNs anyway,
why not go further and restrict 1 LUN to 1 direction only ?

Atsushi Nakamura

BJ Technology Develpoment 22,
Canon Inc.

53 Imai Kami-cho
Nakahara-Ku, Kawasaki-shi
postal no. 211