IFX Mail Archive: RE: Questions to be addressed...

IFX Mail Archive: RE: Questions to be addressed...

RE: Questions to be addressed...

Michael Crawford (mcrawford@iready.com)
Thu, 25 Mar 1999 10:12:08 -0800

> -----Original Message-----
> From: Nick Webb [SMTP:nwebb@auco.com]
> Sent: Wednesday, March 24, 1999 3:00 PM
> To: rshockey@ix.netcom.com
> Cc: mcrawford@iready.com; ifx@pwg.org
> Subject: RE: Questions to be addressed...
-- Snip --

> >Its really not helpful to continue to insist on goals based on the need
> >to support 8 bit controllers. Some vendors are interested in adding
> >value
> (read margins) to their products and if the ultimately agreed to >mandated
> features require more sophisticated hardware components .. so >be it.
> We're talking about networked devices here, and it's unusual to see 8 bit
> devices used in networked office equipment these days. The only one I can
> think of is HBM's print server card which is based on a 64180 derivative.
> We produce protocol stacks for printers, etc, and we've never had a
> customer ask for anything less than a 16 bit machine, and 32 bit machines
> are increasingly becoming the norm.
*** I see what you're saying...I am coming from the world of
consumer devices
not networking. The consumer device IS 8 bit with a few smatterings
of 16 bit.
We aren't even talking about web tv or other high cost products,
we're talking
about toys, the next generation (cheap!) digital cameras, $99 fax
machines, etc.
that are all going to be on the Internet in 2000/2001. What we do
NOW will effect
the cost of these devices as they "get on the net".

Doesn't a printer require a 16 bit or even 32 bit for two reasons?
One reason cuz
the operations required for brief moments in time need to get done
REAL fast,
and second because their code library is not portable and matched
to a particualr
processor? In any case, IPP is for printers and scanners, what I
thought we were
talking about here was for fax. QUALDOCS can be satisified with
full IPP if we
are just talking about quality docs. Print servers are nice, and
can support a lower
capability in IPP (a limited IPP) for that purpose, let's just not
do things that cut
out the 8 bit guys. This is not supposition on my part. If we
aren't careful about
code size and processing horsepower requirements, the QUALDOCS spec
not be used by those folks

> >>> Are you looking at new machines or "black box" add on's?
> >>New machines. Sharp, Brother, Matsushita (Panasonic) have spent
> >snip...
> >>insignificant. Here the BOM cost has to far south of $50.
> >
> >Yea I know .. SRP looks like $149.00 with superstore distribution. I've
> >run these numbers myself. Been there done that.
> Are we considering that a new generation of regular cheap fax machines
> could have this Internet QualDocs mode in them? I had assumed that the
> lowest target machines for this sort of thing were laser printer class
> printers and laser MFPs, the sort of machines that have networking built
> into them today. In these class of machines, OEMs typically pay roughly
> $80
> (and up) to print server card manufacturers to get networking, but in this
> class of machine the incremental cost of QualDocs is negligable.
> Yes, I know we're considering the protocol not a product, but it's good to
> keep sight of the prospective end user...
*** It seems a lot of people are under the mistaken impression that
connectivity is for the MFP which have margins to justify the cost
of adding
the funcionality. MFPs can add new features monthly, and in fact
have done so.

The real issue is the commodity products which need new features to
differentiate them
from all the other low cost commodity products. These devices will
have new
features added only if the cost is low enough and has further room
for cost reduction.
Adding cost by upgrading the processor is prohibitive as is anything
over 200K bytes of
data (actually I might argue anything over 64K bytes gets their
teeth knashing).

These are people who really understand the costs of adding software
both in terms of
development costs as well as what it takes to store the code (ROM
costs), providing
data storage while running (RAM), as well as user perception of
performance (if a device
hangs it is a throwaway, not a service call...and a customer who
never buys again). Think
of a universal TV remote control. If it has quirky operation, not
even considering if it goes away and
seems to be nothing for a few seconds, it will wind up in the trash
can, and SOMEONE ELSE's
universal remote control gets purchased. This is the class of
devices expected to be on the
net in the next two years. The manufacturers are looking at
IPP2IFAX (sorry, but that was
why their interest was peaked!) to help them get capabilities that
the current specs don't help
them with.