Special needs would come under user interface and isn't really anything to do with the communication protocols or physical transports that this thread has been hammering. As long as a PC, smartphone, tablet or whatever can be used as a controller, then it is just ( ) a matter of software in the device to adapt it to the user. Oversimplifying the case, it's a bit like putting a really big knob on a DC controller.
I'm not sure I understand why UDP wouldn't be reliable enough. But then, from my perspective as a Telcom engineer, TCP is almost never the right protocol for a consumer product. #1 problem with the internet if you ask me. As for the expense of smart phones. My motorola droid was $50 from Verizon. Obviously the $30 a month for the data plan is not desirable, but certainly, the hardware itself is cheaper than say a NCE procab. I'm waiting to see if someone comes out with the droid equivalent of an iPad/iPod Touch. As I've said before, part of me really thinks the demand that you NOT use a computer at the behest of a subset of modelers is a mistake for any complex layout communication project going forward. For the types of future looking products we're talking about, I think pushbutton configuration and control will make products cost more and be more complex. In the year 2010, I don't think it's unreasonable to expect people to have and use a computer if they want to control and use complex devices on their layout. Grouse and complain all you want, but the computer will make it simpler. At least that's how I see it.
Aww, c'mon, YoHo. You know better. That droid would cost more like $500 in a fair comparison. The parts count alone in an iPhone is around $175, nevermind the usual overhead. The only way the companies can afford to "sell" you that Droid for $50 is that they're making the other $450 off you from the phone service plan. So, it's not really fair to compare the "plan" cost of a smartphone vs. the retail of a ProCab. Absolutely. Especially since it's brain-dead simple to design a (application specific) computer that looks and acts just like a Zephyr or a ProCab or a DT402 to the user. In fact, just what do those "I don't want a computer on my layout" people think is inside that handheld throttle?
On point 1 The only thing that matters is the real cost to me. If Verizon wishes to subsidize my handset purchase and thus subsidize my DCC throttle purchase, than good for me. The actual cost of the product only matters to the people paying that cost. Doesn't matter at all to me. On Point 2 Exactly. A DCC system IS a computer. Just not a generic one, you can stick your fingers in your ears and scream lalalalalalalala all you want, but it doesn't change the fact that if you have DCC, then you have a computer on your layout. THE HORROR!
I probably shouldn't have brought up the parts cost of the iPhone. It only confused my point. I was trying to underscore that Apple/AT&T (or Moto/Verizon) aren't really selling these things at a loss. My point was that Verizon/Motorola are still getting their $500 out of you, they're just not as "up front" as NCE/Digitrax are. How else do you think Apple posted record profits in the worst recession since the Depression? If Digitrax did the same thing, they could give you the controller for free, as long as you'd be willing to pay a $10/month "LocoNet Fee" for the next two years ... so comparing a "$50" smartphone to a $200 DT402D throttle isn't really fair because the pricing model of the phone hides its true cost to the consumer. LOL! Amen!
I think that $175 is the cost for the iPad including the $100 screen. The iPhone 4 would be less. There are also questions about how and where the iPhoneOS and other overhead is amortized. Whether that's a part of the $175 BOM or not. Still, they're not exactly struggling on profit margin with these. I understand that the business model is different between phones with a contract and DCC, my point is more of a "using your phone as a DCC controller is Cheaper" kind of thing. I already own the phone. I bought it so I could have a phone and text people funny pictures and post on trainboard (please please please get the iPhone skin in the upgrade PLEASE) when I'm not at a computer. If I could also use it as a throttle, then that's just another thing to amortize my costs over.
Oh, of course. I probably should have said that as well. If you consider the phone/throttle as an additional use of something you'd already have anyway, then it really is a "free throttle" 'coz, let's face it, the incremental cost of the throttle usage/functionality is effectively zero, if you're paying for the phone service out of the "telephone" bucket instead of the "trains" bucket. It's only when you say "I want a throttle. Let's go out and buy an iPhone (or Droid)." that the cost comparison fails. In that case, a much more fair comparison would be between the stand-alone throttle and an iPod Touch, which - conveniently - is a wash. Until you realize your DT402 won't play cool background music while you run your trains. Oh, and I had the amazing "luck" of finally giving in and buying an iPhone 3Gs in April . So I don't need a skin. Of course, I don't have that nifty extra camera and FaceTime functon, either.
No no, I don't mean the case for the phone. someone has made a vBulletin Skin for use on smartphones. it's call the iPhone skin for obvious reasons. It cuts down a lot of the features, but makes reading/replying to threads an actual reasonable experience.
I don't think any subset is saying you must not use a computer, it's more that a subset don't want to use a computer as in buy, set up and manage a PC or Mac. Subset also implies a minority, but I suspect that it's the minority that want the complexity that you suggest makes a computer essential - most modellers want to run trains, not become systems engineers. So these standards need to support current and future complex systems BUT they need to support the simple ones as well. Demanding a PC as a master station, for want of a better term, and ownership of a smartphone as a controller will surely leave a huge chunk of the hobby sticking with DCC or DC. Yes, almost everything has a microprocessor in it these days, but there is a big difference between an embedded processor that you plug in and is 'up' in a few seconds and faffing around with loading software onto a PC and waiting for a few minutes every time you turn it on before it does what you want.
But we're not talking about a system for the simple things anyway. If you're interested in, lets say a signalling system and you want it tied into DCC and digital control, then you've already gone super complex. At that point, the PC makes it simpler. In other words, for what we're talking about, people who just want to run trains, don't apply. Lets not get confused about what we're talking about. DC and DCC exist. None of this is going to get rid of that. At least not now. I'm sure BNSF Dispatch would like a "simple" way to control their block occupancy and signalling system too, but they have a massive network with many computers. If you want to mimic on a small scale electronic CTC, then you probably will need a computer. Heck, even people who are doing old school CTC, that's not simple, those giant green CTC boards aren't simple at all. In many ways they're more confusing than a computer. What we're really talking about is people who want the illusion of simplicity. I know people that think that DCC is simple and it would be more complex with a PC. Yet, I'd argue, as an electrical engineer that JMRI is simpler than doing the same on any controller out there. from a cost perspective, removing the PC will raise the cost, because designing a push button interface will cost more money to develop and cost more time to implement and in fact to learn and that will translate to higher MSRPs. So you aren't really going to save money by keeping the computer out of it. Or in other words, there is no logical reason to make it not require a PC. There are emotional and Psychological reasons only.
Sorry, I'd got the impression that these standards were to replace the existing mish-mash of DCC schemes. If they are ONLY aimed at complex systems (and rereading post #1 of the thread does seem to suggest that) then I'd agree that supporting simple systems may not apply. As an electrical engineer I'd disagree a bit. It's fine to say 'these are the standards and here is some software that'll run on a PC to provide configuration and control' but there is no logical reason for the standards to prevent manufacturers producing other interfaces for those that want them - just as they do now in fact. Of course there are emotional and psychological reasons involved - most people do everything in life with a healthy dose of them, which is one reason price/cost is not everything. People will often pay more for something nice - say an iPhone - whereas many others will just buy a cheap mobile phone to make calls with
Let's clarify something. I think it would be a good thing to design the system around a computer, given that an embedded Linux based computer could be built to look/feel like a DT402 pretty easily. Designing the system to require a PC would probably be a bad thing. The good news is that the former allows a PC to be used as he computer, or not. Just wanting to clarify, because different people assume different things when they think of a "computer".
Success! The Wifi-UART module is spectacular... I configured the module to connect to my home Wifi network - Supports authentication and automatic connection. I hardcoded an IP address but also supports DHCP. Basically turn it on and seconds later you can open a TCP Socket to it. Also supports HTTP and UDP. The HTTP support would be interesting as there is the possibility to create a control interface using only a simple web page. So... Interfaced to LocoBuffer, the attached screenshot shows Realterm sending track activate and deactivate commands and the associated datastream received back from the track. The trick now is to write a simple control application via opening a TCP Port number on the iPhone/iPad. Point to note - This would allow anyone today to extend a layout over a Wifi Network using a small piece of software to do virtual com port redirection to TCP. I will write this all up and publish on a new thread. Dwayne
So does this mean I could run JMRI upstairs and use wifi to talk to a loco buffer on my layout downstairs?
OpenLCB is not at all trying to replace DCC, although it will work with DCC. It is aimed at simple systems and novice users -- eg buy two nodes and start using them - programming is by push buttons. It is also aimed at complex systems with multiple heterogeneous segments joined by a higher speed backbone. While a PC will be useful for setup, monitoring, debugging, and even operating, it is not required. As an aside -- studies showed that UDP was not at all reliable, packets were dropped quite frequently - probably for network-stack reasons, perhaps lack of memory. Since TCP/IP is guaranteed reliable, stack writers seem to just drop UDP packets whenever they feel like it.
David Harris wrote "DCC specifications, right or wrong." In the beginning, before the DCC specification, we could run our trains on separate sections of track and if we kept them separate, it worked. Improvements came from such sources as Dynatrol, Keller, CTC 16 and others, each saying 'use mine and you will be fine'. None of these was interested in cooperating with the others as none were compatible. Lenz, in Europe, had a hardened industrial system and in concert with Marklin produced a system much like that now in use. The NMRA in response to members desires developed the DCC standard without too much modification. It was licensed without cost, it works, we can buy hardware that runs together, manufacturers know that what they make will sell. Is it the best it could be, absolutely not! Will making it so that it only runs when supported by a computer; in my mind, not at all! For those of us with fixed layouts it is vital that we use equipment/methods agreeable to the club. I feel that a secondary bus provides the necessary freedom and control; now what approach? Obviously neither 9.5 nor 9.6 to the exclusion of the other is acceptable. Bod
True, very true. It's not likely I will ever see the desired automation, on my layout. Pretty much on a short string financially and I'm on the short end of the stick as time is running out. I can only hope model railroading continues in the direction it has and leaves the old analog DC option, historically the standard of model railroading as is and available. You know, for those of us on the poorer side of the tracks. The difficulty facing NMRA leaders as said by a former president, "The members don't really know what they want"..."There is to much diversification in the ranks". In other words it's difficult to get them focused and working on one project.
Ya all know my feelings on the NMRA. BUT...for those who are new to the hobby...new to TrainBoard...or may feel the same...here ya go ! NOT My Railroad Association Ty...have a nice day .
