Please select Into the mobile phone version | Continue to access the computer ver.
Crash log from Android DJI GO4 v4.1.8
1465 7 2017-9-15
Uploading and Loding Picture ...(0/1)
o(^-^)o
M1dn1ght N1nj4
lvl.4
Flight distance : 11522 ft
United States
Offline

Ive been trying to capture the log of when the APK crashes, and I think I finally got something.  I'm an Android developer, and from what I can tell, there are a LOT of things going wrong in here.  Of course, none that I can fix, as the firmware is not being released under GPL as it's supposed to be.  Just Busybox and a util library.  I can guarentee that almost ALL of the firmware, being based on Android (open source GPL) is under GPL license (minus a couple custom-written libraries from DJI that don't depend on ANY other GPL software).

But for now, here is my most recent crash log:

https://pastebin.com/h6xghAw8

2017-9-15
Use props
Charles Adams
First Officer
Flight distance : 3821312 ft
  • >>>
United States
Offline

My technical days of programming are years behind me, and I've not the skills to evaluate the crash log.  In fact my technical peers and staff typically will use a log analyzer of some kind rather than just eyeballing the log files (though many are sharp enough to do just that).

When you say you see a lot going wrong here...  are you able to derive if it's from the android system itself?  Or if it's within DJI code that everything is "going wrong"?

In particular I wondered about what goes wrong when the app runs on an underpowered system.  I can't figure out coding techniques anymore, but I've thought/guessed that the application performs multithreaded operations, and that an underpowered processor can't keep up with the demands (and thus things go south).  Of course this is just me trying to relive my glory days of coding and debugging, and I've truly no clue.
2017-9-15
Use props
jyc
Captain
Flight distance : 2084534 ft
Canada
Offline

My experience is that 4.1.8 runs like mud compared to previous versions. With some experimentation with antenna orientation I have almost no video artifacts but it still very heavy video latency on my screen.
2017-9-15
Use props
M1dn1ght N1nj4
lvl.4
Flight distance : 11522 ft
United States
Offline

Charles Adams Posted at 2017-9-15 10:57
My technical days of programming are years behind me, and I've not the skills to evaluate the crash log.  In fact my technical peers and staff typically will use a log analyzer of some kind rather than just eyeballing the log files (though many are sharp enough to do just that).

When you say you see a lot going wrong here...  are you able to derive if it's from the android system itself?  Or if it's within DJI code that everything is "going wrong"?

Page after page of error level logs.  I won't go over every single one, but thats WAY too many.  Even a bad commit to a ROM that was pulled from semi-compatible source doesn't give this many errors.

The Pixel XL that I'm running this on is plenty powerful enough to run the app.  The apps API level is what really determines what it will run on.  I'd be willing to bet that the old Galaxy Nexus would run it, although I'm unsure if that phone supports 5GHz WiFi or not.  But remember, that's a REALLY old phone by now.


These are a few of the key things I pulled from the log that set off flags with me:
=================================================================

java.lang.ClassNotFoundException: dji.midware.data.model.P3.DataWifi_gGetPushLog

NetworkProtocol::GetTask::run exception: java.net.UnknownHostException: Unable to resolve host "static.bundles.hybrid.api.here.com": No address associated with hostname

I/WindowManager(1077): Screen frozen for +610ms due to Window{6953c6e u0 dji.go.v4/dji.pilot.main.activity.DJIHubActivity}

E/OMX-VDEC-1080P(864): Extension: OMX.google.android.index.AndroidNativeBufferConsumerUsage not implemented

W/System.err(26307):         at com.android.okhttp.internal.http.StreamAllocation.findHealthyConnection(StreamAllocation.java:126)

E/OMX-VDEC-1080P(864): Does not handle dataspace request
E/OMXNodeInstance(864): getConfig(0xe7c8ffc4:qcom.decoder.avc, ??(0x7f00005e)) ERROR: UnsupportedSetting(0x80001019)

W/GrallocMapperPassthrough(18875): buffer descriptor with invalid usage bits 0x2000





And a million lines of this:

E/NetworkProtocol(26307): NetworkProtocol::GetTask::run exception: java.net.UnknownHostException: Unable to resolve host "static.bundles.hybrid.api.here.com": No address associated with hostname
2017-9-15
Use props
DJI Elektra
Administrator
China
Offline

M1dn1ght N1nj4 Posted at 2017-9-15 19:39
Page after page of error level logs.  I won't go over every single one, but thats WAY too many.  Even a bad commit to a ROM that was pulled from semi-compatible source doesn't give this many errors.

The Pixel XL that I'm running this on is plenty powerful enough to run the app.  The apps API level is what really determines what it will run on.  I'd be willing to bet that the old Galaxy Nexus would run it, although I'm unsure if that phone supports 5GHz WiFi or not.  But remember, that's a REALLY old phone by now.

I would transfer your crash log to our engineers and locate the problem,please wait patiently. Thanks for support.
2017-9-15
Use props
DJI Susan
Administrator

Offline

May I know which type of mobile do you use? What's the OS version?
2017-9-16
Use props
M1dn1ght N1nj4
lvl.4
Flight distance : 11522 ft
United States
Offline

DJI Susan Posted at 2017-9-16 01:31
May I know which type of mobile do you use? What's the OS version?

Pixel XL (Google Play edition), running on Verizon, 32GB
2017-9-16
Use props
DJI Susan
Administrator

Offline

M1dn1ght N1nj4 Posted at 2017-9-16 06:34
Pixel XL (Google Play edition), running on Verizon, 32GB

I understand your frustration, but the stability can not be ensured for mobile devices that have not been tested. I will forward to our engineers as well and hopefully they will optimize in the future.
2017-9-20
Use props
Advanced
You need to log in before you can reply Login | Register now

Credit Rules