Betta release is not stable. The application freezes on android under different conditions:
Sometimes when entering the garage
Sometimes when entering can monitor
I am trying to write a debug file and send it to the server. at the moment I can not provide more information on the serial port, since the application crashes. it helps only restart the android. Android 7.1
I sent a 10kb log file to the server. I hope it will be useful.
I reset the cache and application settings. So far, everything seems to be fine. The application is stable.
I don’t see much difference in the work of the serial port.
All the same when sending a large number of frames through RealDashCAN-arduino. After a while, the data freezes. If the stream is not very full, then the data is smart and not interrupted.
There are no special changes for me.
We can say that everything works as before.
Most likely the matter is in my sketch, or arduino.
Sometimes the update is unstable until you clear the data or uninstall/install. I do not know if that is something related to RealDash, or Googles system that does not download the entire package when updating. 
Seems like my fix had no effect on the problem you have had. But at least it did not go backwards, the CAN frame parsing should be much more stable now in case of corrupted incoming stream.
OK. This is not noticeable in the eye. But if the analysis is better, then this is only to the benefit. But I began to notice periodic hangs of testimony.
I do not understand what it is connected with, I do not understand how this is manifested. It’s just that the data freezes and it’s not possible to influence the application. Only reset through the application menu and re-entry.
Then everything works stably. Until I found a pattern. And returned to the stable release.
What does this only happen when the RealDashCan connection is connected. On my second test device, without RDCAN, I do not observe this.
I will check Play store crash/hang logs if I can find something there.
Can’t see anything on Play store crash logs. I had two Android devices running RDCan protocol for all night with zero problems so the hangs you are experiencing are probably caused by something else.
OK. I will watch at home. Thank!
I found something that may have caused the hang in 1.7.4-BETA. Will push a new Beta later today.
I just wanted to write that I skated 173 all day fine, as soon as I installed 174, it hangs again: lol:
In general, we are waiting)))
3 hours allocated for tests 174.b.3 …
The result is positive. Not a single glitch.
We observe further. If there are failures - I will inform.
What was the reason?
Great, thanks for testing! The RD CAN data parser did hang if data was coming in slowly one byte at the time and the tag was not found on 14 byte boundary.
Hmm … Do you mean one byte per program cycle? Is this even possible?
If by default the program sends 8 bytes of only useful information in the frame.
Or did you specifically simulate slow sending?
It is simply brilliant. Your correction corrected all the flaws about which I generally ever wrote on the CAN protocol.
No friezes.
No overlay.
No trash.
No shutdown.
No freezes.
Now you can read the full set of packages automatically, without selecting individual packages.
T.E. Any set of packets will be read on any vehicle with a CAN ceiling. (subject to normal iron)
No lags, no brakes.
174.b.3 - in the form as it is, the most stable of all stable ones.
Thank you
I did simulate the slowdown with 50% added invalid bytes and thats how I found the problem.
Its not typical, but can happen, especially on Bluetooth connection where I have seen behavior like getting number of bytes in like this:
52-37-53-74-1-1-1-1-1-5-3-67-45 etc.