|
Cedar Rapids, Iowa | Good stuff Matt. Thanks.
Counterpoints:
1: Point taken. Has there been any testing of 100Meg twisted pair Ethernet in these environments to see if it is susceptible to any potential electrical noise?
2: That’s great, but does it really matter on this network? You send a message, maybe retry once. If you don’t get an acknowledge back within 100ms, alert the user and go into fail safe mode.
3: Back when people used Ethernet hubs, collisions were a problem. Now that everyone uses switches, packets are buffered and forwarded. I am aware that a saturated network could potentially block some messages from getting though, but that would require a 100Meg network being saturated. If you have 2 computers on a LAN at your house, and they are moving a massive file, you can still surf the Internet.
4: Really? I didn’t know these devices were intelligent enough to do message filtering at the hardware level. Does the chip keep track of the name:address translation (even if the peer gets a new address after being bumped by another device).
4b: You seem like a very knowledgeable individual, so I’m sure you’re aware Ethernet functions the same way with message filtering at the hardware level. Only packets with the device’s MAC address or the broadcast address are sent to the software for processing.
4c: I would love to know what you consider “low cost” for a device that manages the tail lights on a vehicle. Development time costs money, and this stuff isn’t produced in any quantity that allows that cost to be divided across a ton of shipped units.
While the large manufacturers appear to be pushing forward, it seems most of the shortline manufacturers are in a bit of a quandary. In order to develop a product that works on ISOBUS, they’re going to have to either find talent that already knows how the standard works, or pay some people to learn it. Either way, this is complex enough that it is going to cost time/money. When you have to break that price down over a small volume of units, how much extra $ do you have to charge for each to pay for the development? Are the customers willing to pay for it?
When you say that you foresee an additional network alongside ISOBUS, what type of traffic would you expect to see on each network?
When I go against the grain on something, I have to wonder if I’m being unreasonable. Maybe that’s the teenager in me that “knows it all”. It just seems to me like trying to get communication between devices with ISOBUS is like reinventing the wheel, and ending up with an octagon. Yeah, it kind of rolls, but at the end of the day, all the customers look at it and wonder how you ended up with an octagon when the rest of the world has been using round wheels for years.
-Lance | |
|
ISOBUS – thoughts from a geek
