Meindert,
Very nice to hear from you again. You have been away quite some time. I
can't believe I am hearing this from you. You are the perfect person for
this thread. I think you need to think a bit outside of the box. As you
know, each NMEA manufacturer today is addressing the inadequacies of NMEA
with their own propriety solutions and selling them as the next best thing
in boat electronics, like SeaTalk. Yet we have a huge, inexpensive
commercial infrastructure all around TCP/IP and yet the marine industry is
trying to reinvent the wheel. You should revel in this foolishness and
consider this as a golden opportunity to develop a transport network like
the CAN bus SAE J1939 standard, but using TCP/IP as the flexible transport
medium. Where the entry and exit ports are box standard NMEA, but are in
fact intelligent gateways to the Ethernet transport. You can buy off the
shelf single chip TCP/IP support and inexpensive switches. I see these
gateways programmable as talkers or listeners with a central
router/controller accepting the NMEA inputs and buffering them as well as
distributing them by IP address at any rate the listener required. This
solution solves all the NMEA problems and by developing additional gateway
flavors, solves all the compatibility issues between devices and
manufacturers. Most of this already exists inexpensively. All it takes is a
little ingenuity to integrate it into a total package. I think the market is
huge. There are a lot of floating customers out their just waiting for this.
Please also keep in mind that this same transport can also move all data
types including other, unrelated traffic like audio, video and other
computer related data streams.
Steve


"Meindert Sprang" wrote in message
...
"Bill Kearney" wrote in message
t...
He's also failing to grasp the TINY size of the marine electronics
market.
Much like the naive fools that rant about how their boat isn't serviced
like
their Honda.

And that is exactly why marine instruments will not support an ethernet
interface with TCP/IP because it is simply too expensive to implement. And
surely people will now tell me that I can buy an ethernet card for my PC
for
less than $5. But this will simply not happen for the relatively small
marine market.

Meindert

Hi Steve,

"Steve Lusardi" wrote in message
...
Meindert,
Very nice to hear from you again. You have been away quite some time.

I'm lurking here every day... so not really away :-)

I can't believe I am hearing this from you. You are the perfect person for
this thread. I think you need to think a bit outside of the box. As you
know, each NMEA manufacturer today is addressing the inadequacies of NMEA
with their own propriety solutions and selling them as the next best thing
in boat electronics, like SeaTalk. Yet we have a huge, inexpensive
commercial infrastructure all around TCP/IP and yet the marine industry is
trying to reinvent the wheel.

Well, I think it is not that simple. Off course we have thousands of cheap
products for ethernet networking. Most of which are not suitable nor allowed
in marine environments. Take the average UTP CAT5 cable: not permitted on
board of SOLAS vessels. The average hub is not IEC945 compliant: not
permitted on SOLAS vessels. Not to mention the average RJ45 connector...

Furthermore, while everyone is hammering on using TCP/IP to replace NMEA:
TCP/IP is the least suitable protocol for this. In a marine network, one has
several devices all sending information to whoever it concerns. TCP/IP on
the other hand, is a point to point protocol. UDP broadcasts would be much
better since they reach every device on the network. Look at the average
Serial-Ethernet bridge: they all to TCP/IP to replace ONE serial link. Not
suitable. Look at the price of these little boxes compared to bog standard
ethernet cards and you see how in a relatively small marine market prices
would increase when you equip devices with an ethernet interface.

You should revel in this foolishness and
consider this as a golden opportunity to develop a transport network like
the CAN bus SAE J1939 standard,

NMEA2000 is based on CAN

but using TCP/IP as the flexible transport
medium.

Do you realise that basic CAN only transports 8 bytes per packet at a time?
To put TCP/IP on top of that causes a huge overhead on the network, not to
mention the burden on the processor that drives the CAN controller. CAN was
never invented for this. CAN was invented to broadcast data on a network to
every one who needs it. No point to point connections. CAN is perfect for
distributing navigation info.


Where the entry and exit ports are box standard NMEA, but are in
fact intelligent gateways to the Ethernet transport. You can buy off the
shelf single chip TCP/IP support

At a price....

and inexpensive switches. I see these
gateways programmable as talkers or listeners with a central
router/controller accepting the NMEA inputs and buffering them as well as
distributing them by IP address at any rate the listener required.

The speed of NMEA is so low that you can simply dump it on an ethernet
network as it comes, without any intelligent distributing or rate control.
Do some math: 100Mbit/s vs 38400 b/s: That is the equivalent of 2600 AIS
receivers spitting out data continuously one one UTP cable.

Meindert

"Steve Lusardi" wrote in message
...
Meindert,
Very nice to hear from you again. You have been away quite some time. I
can't believe I am hearing this from you. You are the perfect person for
this thread. I think you need to think a bit outside of the box. As you
know, each NMEA manufacturer today is addressing the inadequacies of NMEA
with their own propriety solutions and selling them as the next best thing
in boat electronics, like SeaTalk

WERE reinventing, past tense. NMEA2000 is the solution for it, and it works
QUITE well.

Yet we have a huge, inexpensive commercial infrastructure all around TCP/IP
and yet the marine industry is trying to reinvent the wheel. You should
revel in this foolishness and consider this as a golden opportunity to
develop a transport network like the CAN bus SAE J1939 standard, but using
TCP/IP as the flexible transport medium.

Which screams of how little you understand about instrumentation networks.

I think the market is huge. There are a lot of floating customers out
their just waiting for this.

I call bull****. List actual numbers, not pie-in-the sky hopes.

Please also keep in mind that this same transport can also move all data
types including other, unrelated traffic like audio, video and other
computer related data streams.

Which, again, screams of how little grasp you have of how instrumentation
networks function.

Bill,
The problem is across the entire marine spectrum, not just pleasure craft.
There are thousands of commercial vessels that not only bus the nav gear
through NMEA and other IMO approved interfaces,but now also host Ethernet
networks as well. The IMO is a very conservative and at times very backward
organization. I do not agree with Meindert, but he does raise very valid
points. NMEA 2k is better than 0183, but it doesn't hold a candle in
transport capability or flexibility in comparison to Ethernet. It is no
longer necessary nor desirable to host stove pipe transports for different
purposes. The world has changed. I am an electronic engineer that has been
involved with both IT and aircraft instrumentation for 40 years. the world
has changed, we need to keep up. Ethernet and TCP/IP is used by billions
world wide. Implementing this technology allows this "very small" market
place you speak about enjoy the cost advantage of a technology used by the
world.
Steve

"Bill Kearney" wrote in message
t...

"Steve Lusardi" wrote in message
...
Meindert,
Very nice to hear from you again. You have been away quite some time. I
can't believe I am hearing this from you. You are the perfect person for
this thread. I think you need to think a bit outside of the box. As you
know, each NMEA manufacturer today is addressing the inadequacies of NMEA
with their own propriety solutions and selling them as the next best
thing in boat electronics, like SeaTalk

WERE reinventing, past tense. NMEA2000 is the solution for it, and it
works QUITE well.

Yet we have a huge, inexpensive commercial infrastructure all around
TCP/IP and yet the marine industry is trying to reinvent the wheel. You
should revel in this foolishness and consider this as a golden opportunity
to develop a transport network like the CAN bus SAE J1939 standard, but
using TCP/IP as the flexible transport medium.

Which screams of how little you understand about instrumentation networks.

I think the market is huge. There are a lot of floating customers out
their just waiting for this.

I call bull****. List actual numbers, not pie-in-the sky hopes.

Please also keep in mind that this same transport can also move all data
types including other, unrelated traffic like audio, video and other
computer related data streams.

Which, again, screams of how little grasp you have of how instrumentation
networks function.

The problem is across the entire marine spectrum, not just pleasure craft.
There are thousands of commercial vessels that not only bus the nav gear
through NMEA and other IMO approved interfaces,but now also host Ethernet
networks as well.

Thousands does not equal economies of scale typical for computer electronics
markets.

That and, iirc, ethernet has no standardized connectors for watertight
fittings. Then there's the hassle of all the wiring having to be home-run
back to a switch. There's no way to daisy-chain the instruments along a
single backbone. So it's more wire to break, more connectors to leak. No
thanks.

The IMO is a very conservative and at times very backward
organization. I do not agree with Meindert, but he does raise very valid
points. NMEA 2k is better than 0183, but it doesn't hold a candle in
transport capability or flexibility in comparison to Ethernet.

And what capability and flexibility claims are so great as to be useful in
the MARINE industry? Just what about TCP/IP is so useful in this
application?

The world has changed. I am an electronic engineer that has been involved
with both IT and aircraft instrumentation for 40 years. the world has
changed, we need to keep up.

NMEA2K keeps up, and more.

Ethernet and TCP/IP is used by billions world wide. Implementing this
technology allows this "very small" market place you speak about enjoy the
cost advantage of a technology used by the world.

How, exactly? More wire, non-standard connectors (RJ45 in a screw cap?
puh-leeze)

I'm all for cost effective solutions. But, as the saying goes, when all you
have is a hammer, everything looks like a nail.

Something I'd like to point out: The market goes beyond boating. There
are others including aviation, automotive, sporting, surveying, research,
and I'm sure many others. NMEA is present in most if not in all of these
areas.

This is another reason why I'd like to see changes made. Open architecture
that is extensible, so that these other areas may be addressed easier.

The electrical part of the standard is brilliant in the use of CAN. My
argument is mainly about the sentence structure and how it is retreived and
used. 0183 is great because it is human readable, somewhat rs-232
compatible, and easy to implement. Also there is enough information
floating around the web to figure out how to use it. This is not the case
with 2k.

All this makes it easy to "plug" into a laptop and test or use the talkers.
---- Posted via Pronews.com - Premium Corporate Usenet News Provider ----
http://www.pronews.com offers corporate packages that have access to 100,000+ newsgroups

"Poit" wrote in message
00.119...
Something I'd like to point out: The market goes beyond boating. There
are others including aviation, automotive, sporting, surveying, research,
and I'm sure many others. NMEA is present in most if not in all of these
areas.

Indeed, because it's cheap and easy to implement.
Anything else will be more expensive and more of a hassle to connect.

This is another reason why I'd like to see changes made. Open
architecture
that is extensible, so that these other areas may be addressed easier.

Could you please explain where NMEA 0183 fails in this respect?
It is extensible, NMEA 0183 allows for "Proprietary sentences" which can be
arbitrarily defined and it really is an open standard.

Also there is enough information floating around the web to figure out
how to use it.

Ah, so that is your real point: you want all the information for free. You
think because of the fact that you have to pay to get the information, the
standard is not open. Wrong. NMEA is an open standard and available to
anyone. A closed standard like Seatalk is NOT available, except for the
reverse engineered stuff on the web.

Meindert

Something I'd like to point out: The market goes beyond boating. There
are others including aviation,

I'm sure the FAA would laugh in your face at that.

automotive, sporting, surveying, research,
and I'm sure many others. NMEA is present in most if not in all of these
areas.

And working quite well. Besides trying to shore up a weak argument, is
there a point here?

This is another reason why I'd like to see changes made. Open
architecture
that is extensible, so that these other areas may be addressed easier.

Bull**** again, and Meindert calls you on it in the next message.

The electrical part of the standard is brilliant in the use of CAN. My
argument is mainly about the sentence structure and how it is retreived
and
used. 0183 is great because it is human readable, somewhat rs-232
compatible, and easy to implement. Also there is enough information
floating around the web to figure out how to use it. This is not the case
with 2k.

Ah yes, the old "human readable" bogus argument. Are jpeg files human
readable? How about mp3 files? And yet they're amazingly useful in meeting
the needs of their applications.

"Bill Kearney" wrote in
t:

Something I'd like to point out: The market goes beyond boating.
There are others including aviation,

I'm sure the FAA would laugh in your face at that.

Why would they do that? Besides ARINC, NMEA is a format found there.

automotive, sporting, surveying, research,
and I'm sure many others. NMEA is present in most if not in all of
these areas.

And working quite well. Besides trying to shore up a weak argument,
is there a point here?

The point is that you are trying to make it seem that the market is too
small for changes, when in fact it goes well beyond your little world.



This is another reason why I'd like to see changes made. Open
architecture
that is extensible, so that these other areas may be addressed
easier.

Bull**** again, and Meindert calls you on it in the next message.


I know about the proprietary sentences. What is your
major malfunction?

The electrical part of the standard is brilliant in the use of CAN.
My argument is mainly about the sentence structure and how it is
retreived and
used. 0183 is great because it is human readable, somewhat rs-232
compatible, and easy to implement. Also there is enough information
floating around the web to figure out how to use it. This is not the
case with 2k.

Ah yes, the old "human readable" bogus argument. Are jpeg files human
readable? How about mp3 files? And yet they're amazingly useful in
meeting the needs of their applications.



What"s bogus about that argument? Jpeg and mp3 standards are readily
available on the net. Also I'm not arguing that NMEA does'nt do it's
job. In fact I was standing up for NMEA 0183. What's wrong with having
it human readable? If something breaks having it easy to work with makes
troubleshooting a lot easier.

I don't understand why you keep missing the main point. Maybe you work
for NMEA? You'd think that by the very nature and even the title of this
newsgroup, the audience would be in favor of the do-it yourself. Even
some of the other posters see the point.
---- Posted via Pronews.com - Premium Corporate Usenet News Provider ----
http://www.pronews.com offers corporate packages that have access to 100,000+ newsgroups