DCS app bug?

There are many bugs in the DCS app that are known. Some are minor while others can cause an engine to runaway at full speed. For example, there is a latency issue communicating with the TIU that causes wild voltage swings and freeze-ups using variable voltage for conventional control. And there is also the known issue of having the + button almost touching the 200 mph selection, which frequently causes most accidental runaways. Despite knowing of these issues, MTH hasn’t updated their app in several years.

Don’t expect anyone to fix this app. It has never made money. There are no subscriptions to generate revenue. Everyone who wants this app has already paid for their perpetual license. This app has always been a loss leader supported only by ongoing sales of new DCS locomotives by MTH. I have heard vague talk from MTH to the effect that a group of former employees may take over this technology. No firm arrangements have been disclosed. Bugs are the smaller problem for this app. Without ongoing support, the next iOS update could easily break the app with no warning making it useless. And yet here we are with MTH just getting ready to sell their new wireless TIU that relies on this app that nobody is going to be around to keep running. Maybe I am wrong, but again there has been no firm commitment from MTH or any successor to support this app in the future.

I’ve seen all those bugs.  Even shot a video of the needle moving on its own and sending it to MTH.   I’ve seen way to many glitches to list here, one of many reasons I don’t use the app anymore.

And folks wonder why I'm stockpiling DCS remotes and TIU's.

I had a similar issue, but in my case the needele kept returning to 0. Following advice from onr of the members of thsi forum, I su=imple deleted the app and reinstalled it. The problem then went away.

@gunrunnerjohn posted:

And folks wonder why I'm stockpiling DCS remotes and TIU's.

LOL John, don't take them all!

I have a DCS remote on pre-order from the last production run which should arrive in the coming months. I’ve never used a DCS remote but now it seems like a good idea to have one in order to maintain operation into the future without the unease of depending on an app that could implode with no notice.

@mike g. posted:

LOL John, don't take them all!

All I can find at a reasonable price.  However, I think "reasonable price" has gone out the window now, so I'll have to live with my existing stock.

I haven't seen this specific issue.  One thing I've noticed is that the quality of the WiFi connection really affects latency quite a bit.  I found that putting a wireless AP in the basement near my layout vastly improved the the communication between the the smart device and WIU.  I think the WIU can also be hardwired to the network with the Ethernet port.  This would reduce the latency considerably since only the smart device would be communicating over WiFi.

I've sent bug reports to MTH for the DCS app and they have responded, so I think they will update the app in the future.

I haven't looked into it in detail, but since the traffic goes over WiFi, it should be trivial to do a packet capture and determine how the app communicates to the WIU and then "roll our own" app if the future MTH or DCS company drops the ball.

A smart move would be to just open-source the code, and then the computer programming savvy folks like myself could just fix the bugs for them.

@rplst8 posted:

I haven't looked into it in detail, but since the traffic goes over WiFi, it should be trivial to do a packet capture and determine how the app communicates to the WIU and then "roll our own" app if the future MTH or DCS company drops the ball.

Trivial would not be how I'd define the task of creating a new remote package!

@gunrunnerjohn posted:

Trivial would not be how I'd define the task of creating a new remote package!

I didn't say that it would be trivial to create a new remote. 

@rplst8 posted:

I didn't say that it would be trivial to create a new remote.

No, but you certainly alluded to it.   It's pretty easy to misunderstand what you're trying to say here.  The fact that it might be trivial to map the packets is moot unless you're going to do something with that information.

@rplst8 posted:

I haven't looked into it in detail, but since the traffic goes over WiFi, it should be trivial to do a packet capture and determine how the app communicates to the WIU and then "roll our own" app if the future MTH or DCS company drops the ball.

@gunrunnerjohn posted:

No, but you certainly alluded to it.   It's pretty easy to misunderstand what you're trying to say here.  The fact that it might be trivial to map the packets is moot unless you're going to do something with that information.

This thread was about the app, not the remote.  I said it would be trivial to do a packet capture, decompose the protocol, and then make our own app.  Similar to how Mark DiVecchi reverse engineered the sound file format.  Would it take effort? Yes. Would it take some time? Yes. Would it be hard? Not really.

There are a few things that could make it troublesome, but I’m cautiously hopeful that MTH wouldn’t have taken to draconian measures like encryption to mask the function of the communications.  

@rplst8 posted:

This thread was about the app, not the remote.  I said it would be trivial to do a packet capture, decompose the protocol, and then make our own app.  Similar to how Mark DiVecchi reverse engineered the sound file format.  Would it take effort? Yes. Would it take some time? Yes. Would it be hard? Not really.

There are a few things that could make it troublesome, but I’m cautiously hopeful that MTH wouldn’t have taken to draconian measures like encryption to mask the function of the communications.  

I would hesistate to trivialize the effort required to "roll" your own app. Let's assume you have already done your magic and decompiled the iOS app, or "decomposed" the protocol, etc. Even better, let's assume someone handed you the entire DCS protocol and all that was required was some coding. Is it really so easy? It might instructive to review the work of those who have walked this path before you.

First, let's begin with MTH.  They are the experts. They invented the whole thing. Yet, they haven't updated the DCS app in at least 2 years according to the version history. Apparently, it's too much of a PITA, too time consuming, or simply too expensive for MTH to keep up with their own flagship control system.

Secondly, let's look on the Lionel side of app development.  Despite having a known and lengthy list of bugs to fix, Lionel's iCab hasn't been updated in over 4 years.  Also consider the current state of the Lionel LCS partner program. They have released the Legacy protocol. It's right there for anyone to have. However, apparently the few takers that have released apps don't find it that easy to develop or maintain their work. Visit the LCS parter program website and you'll find three broken links to the 3 developers that tried.  I have the "High Rail" app. It's great, but the developer gave up half way along and never implemented TMCC control or many other features. The other developers seem to have disappeared entirely.

The take-away from reviewing the work of others seems to be that these apps are a tremendous effort and time investment.

@rplst8 posted:

This thread was about the app, not the remote.  I said it would be trivial to do a packet capture, decompose the protocol, and then make our own app.  Similar to how Mark DiVecchi reverse engineered the sound file format.  Would it take effort? Yes. Would it take some time? Yes. Would it be hard? Not really.

You might also talk to the Hikel brothers, they were intimately involved in the MTH app development in the early days.  I'm fairly sure they'd dispute that it would be "trivial" to develop your own app.  That's especially true as they were handed the protocol and assistance from MTH, you'd be moving on without any of that assistance.

Look, I'm not saying that it can't be done, I'm just saying it's not trivial.

Greg,

I respect your analysis, and John's.  Based on the information you cite your logic is very sound, but it is limited information at best.  I think that both MTH and Lionel are having trouble supporting the apps not because it's hard to do so, but because it's hard to find people who know how to do so.  These are two different things.

There is much money to be made in smartphone apps.   Big opportunities.  Although this money comes one small sale at a time there are literally billions of these devices in use.  And, there are new apps being introduced constantly.

The challenge for MTH and Lionel is to find software folks willing to trade those big opportunities for our little ones.

Having been involved with several big-opportunity apps over the last few years I believe that there's a shortage of people instead.  I'm with rplst8.  This is not that difficult technically.

It's a business problem.

M.H.M.

I look forward to your new DCS application, glad it's easy.

@GregR posted:

I would hesistate to trivialize the effort required to "roll" your own app. Let's assume you have already done your magic and decompiled the iOS app, or "decomposed" the protocol, etc. Even better, let's assume someone handed you the entire DCS protocol and all that was required was some coding. Is it really so easy? It might instructive to review the work of those who have walked this path before you.

i never said the whole process is trivial. I said it would be trivial to capture the signal over the network.  I also never said ANYTHING about decompiling the iOS app.

Since the app uses WiFi, everything in flight over the air is encoded in Ethernet frames.  If you connect it to home WiFi network, you’re able to see all the network traffic.  Once you have that you can see what the payload is on the Ethernet frames.  It will very likely use TCP/IP, so setting a machine with Winpcap and Wireshark, and setting the network adapter in promiscuous mode will allow you to see all the communication between the app and the WIU.

Then it’s just a matter of hitting buttons on the app, and recording the packets that are sent.

First, let's begin with MTH.  They are the experts. They invented the whole thing. Yet, they haven't updated the DCS app in at least 2 years according to the version history. Apparently, it's too much of a PITA, too time consuming, or simply too expensive for MTH to keep up with their own flagship control system.

I don’t know for sure, but my guess is they subcontracted the work out to an app developer, and the contract has run out on that.  Again, no info but just a hunch.

@GregR posted:

Secondly, let's look on the Lionel side of app development.

I have not seen or used nor do I know anything about the Lionel app.  I never made any claims about it.  

@Mellow Hudson Mike posted:

The challenge for MTH and Lionel is to find software folks willing to trade those big opportunities for our little ones.

Having been involved with several big-opportunity apps over the last few years I believe that there's a shortage of people instead.  I'm with rplst8.  This is not that difficult technically.

It's a business problem.

M.H.M.

Absolutely.  When you look at starting salaries and perks out west at the big tech companies for software engineers and programmers, it can be very hard to find IT talent for smaller companies.

That said, if I thought I could make a decent living at it, I would write apps to control trains all day!

I just started using the app, getting into command control for the first time, and came out with a fairly large number of bugs:

-On ipad the horn button doesnt work for any legacy engine

-On android, legacy lashups crash the app when changing speed

-On android, DCS lashups continually error out

I'm not really sure who to contact from MTH to report bugs to but I'm a software developer for one of those large software companies out west and I wouldn't mind fixing the app for free if that was an option.

@DashingDanLIRR posted:

-On ipad the horn button doesnt work for any legacy engine

FWIW the horn / whistle works fine on my iPad.

@MartyE sorry didnt realize thats a legacy engine.

For me I was testing with legacy diesels and it wasnt working.

What ipad version are you using? and what IOS version?

Thanks,

Dan

It’s definitely an older iPad running older iOS, not at home at the moment, but I’ll check when I get back.

Interesting I tried to connect both my iPhone and iPad just now and can’t get either to connect. Last time I used it worked fine. Guess I’ll be doing some troubleshooting this evening.

Never fails. Even went back to the MTH mode for WiFi.

i am new to this.is DCS problematic.I read one version isnt compatable with the new version.I am staying away because of ignorance.the few mth engines I have are run conventionally.TMCC/legacy is simple.then I hear about star wiring.I am confused.    I just know MTH makes some beautiful stuff

The Lionel Ipad app works well, it's missing some features the cab2 has but I didn't encounter any errors when using it with the LCS wifi module.

I'm happy to help the MTH guys fix their code for free just don't know how to get in contact with whomever owns the app now from their side