Larry,
Actually, I think exactly the opposite is true. I believe NMEA multiplexers could be made using NMEA over TCP/IP on an Ethernet
backbone transparently to the end NMEA device. This technique is used commonly in existing networks today, where the TCP/IP header
is used exclusively for routing purposes and at the endpoint device, the TCP/IP header is stripped and the NMEA sentence is
presented to the NMEA customer bit serial transparently in the normal manner. Similar devices already exist on the market as
Ethernet gateways to RS232/422 devices. Additionally, advanced QOS is available for priority routing if required, preserving the
advantage of manufacturer independence and all the advantages of Ethernet. I personally believe this represents a golden
opportunity.
Steve

"Larry" wrote in message ...
"Steve Lusardi" wrote in
:

Ethernet has it all covered, but Meidert and I have had this
discussion a while back and I didn't convince him then. Maybe today,
he's older now. Steve


Meindert, no offense, but he makes NMEA multiplexers. Ethernet would be
the end of that business.



--
iPhone 4 is to cellular technology what the Titanic is to cruise ships.

Larry

"Larry" wrote in message
...
"Steve Lusardi" wrote in
:

Ethernet has it all covered, but Meidert and I have had this
discussion a while back and I didn't convince him then. Maybe today,
he's older now. Steve


Meindert, no offense, but he makes NMEA multiplexers. Ethernet would be
the end of that business.

Not at all. I'll be having a MiniPlex-2E in a couple of days.

Meindert

"Steve Lusardi" wrote in message
...
Larry,
You normally have your act together, but not this time. Meindert is
correct about his stated risks, but he is overstating the cost
of Ethernet solutions. Today they are canned in firmware and reasonably
priced. The savings in installation complexity,
drastically increased speed and system flexibility far exceeds any
increase in component cost. Please don't get me wrong. 802.11
has its place, but not in systems where data integrity are safety issues.
CAN solutions work well and meet equipment
manufacturer's requirements, but as an end user I must counter with a
standardization argument. The only thing that is standard is
the CAN pipe, everything else is proprietary and manufacturer unique. The
ability to mix and match instruments and devices from
different manufacturers would virtually disappear. Furthermore, the CAN
bus is slow in relation to Ethernet, reducing the amount
of net users and traffic the pipe could ultimately handle. Then there is
the question of physical pipe length and noise
susceptibility. Ethernet has it all covered, but Meidert and I have had
this discussion a while back and I didn't convince him
then. Maybe today, he's older now.

And you still cannot convince me of the idea that an ethernet interface can
be as cheap as an RS-422 interface. The canned solutions you talk about
still cost considerably more than a simple Rs-422 tranceiver chip. Of course
I could have taken an ethernet version of the controller I use in my
multiplexers, but I still would have to add a transformer and at minimum a
UDP stack and, if you want the ease of ethernet, the complete DCHP stuff to
the existing software. You simply cannot compare a dedicated non-mass market
device to mass market PC stuff which cost next to nothing for two purposes:
1) a huge market and 2) a complete multi megabyte OS to support the dumb and
cheap hardware.

This is the same discussion as with Bluetooth. You can buy BT dongels for a
couple of dollars/euros but these rely on Windows to operate. The modules I
use cost 20 times that much because 1) the volume is much lower and 2) they
contain a complete processing system to do what windows would otherwise have
done.

Meindert