Posts by SoftWarewolf

Quote from Esbol

Hi. It would be nice if there was an extension that checks if the app is installed on the phone. For example: you can reward the user for installing the app. Go to the application if installed

I made a little extension for this before
Please login to see this link.
thread: Please login to see this link.

I made a little extension for this before
Please login to see this link.
thread: Please login to see this link.

I published a new app update with the else block added same as your screenshot I believe, and it doesn’t crash anymore.
I don’t know what the user sees tho, I guess for cases where it crashed before, it just doesn’t show the image.

Hopefully Fernando can find the source of the problem that i suspect is with the active picture object/downloaded images, for now I just put the else { block back in, which should help it not crash if for whatever reason the file is missing.
You may be able to do it as well without having to wait for an update if you know how to modify the java files in the runtime .zip.
In CRunApp.java you can add else { on line 1998 after the IF block, then close the else bracket with } around line 2030, this skips the whole portion of code crashing by a null pointer exception if the file is null.

I am also seeing crashes at (CRunApp.java:1999) with the latest update, it looks like the else block after checking if file is null got removed, so it must be trying to get the hashcode of null in some cases causing this crash.
I have not reproduced this on my end, but can see it on the play console. I found the change comparing the code in crunapp.java between a previous version and the latest version.
Dobermann, do you use active picture to load images from url in your app? I am just suspecting that could potentially be the initial cause.

I seem to be getting a correct unique string for each device, but looking at the code this identifier is retrieved with TelephonyManager getImei() correct?
I thought this identifier was heavily protected, requiring READ_PRIVILEGED_PHONE_STATE etc. In these privacy focused times, I am guessing it's bad practice to rely on this identifier and curious how the android object is always able to retrieve this identifier without the need to request any permission to do so?

Quote from haoujey

Hi SoftWarewolf,
How do you get the build files ?

Thanks

This basically only works when the build fails, you could just add a spelling mistake in the source of any android extension you use (unzip, use any text editor to remove a ; or something like that and zip it back up)
It's built in C:\Users\*you*\AppData\Local\Temp in a folder called And###.tmp (### = some number, just search by date modified), as long as the error message that the build failed stays open the files stay there, as soon as you close the message or the build completes then the files are removed. You can copy out this folder and load it up in Android Studio. (just remember to fix the extension you changed, and remove the delete root part of the build.gradle)
After having your project in android studio, you can hold shift while exporting from fusion to export raw files including the application.ccn which you can copy to \app\src\main\res\raw\ to update the app (sometimes fails to work if you've added additional kinds of events or any new extensions etc.)

I have been using AndroidX for a long time on fusion built apps already, as i grab the build files and load my projects into Android Studio.
gradle.properties has android.useAndroidX=true
My last build.gradle had these implementations and others:
'androidx.multidex:multidex:2.0.1'
'androidx.core:core:1.3.2'
'androidx.appcompat:appcompat:1.2.0'
'androidx.legacy:legacy-support-v4:1.0.0'

Thank you for the quick response, I've should have just checked if there was a new beta version available.
292.27 is working well.

I noticed 4.2 (17) support is missing on these later updates, when i tried to enable and launch it on 4.2 emulator, it immediately crashes with "libc++_shared.so" not found. I guess it's beyond time to phase this version out, but i want to update an app that has active users on Android 4.2, so it would be nice if it could somehow still be compatible.

It looks like whenever the main activity is paused and resumed, some black anti-aliasing is added to strings already displayed. Which makes them look a bit shrunk and not matching up with strings created or modified afterwards.
This didn't happen previously, but does in update 292.26.
If you have a fix please do send me the .java files to replace, or just let me know what to change on which line number.

.mfa, .apk and screenshot:
Please login to see this link.

There was a major sound issue in the last beta i tried, sound at max volume with no way to mute or reduce it, volume not following the media volume anymore.

Yeah makes sense, and the first request takes much longer than subsequent requests, so the cache does work to reduce the time after the first request, but it's always a lot slower than http.

Kind of related, I did some testing with https and it seems to take about twice as long as a http request, both on windows and android with the get object, and both using an apache server and an nginx server. Is https really this slow for everyone or could a modification to the Get object potentially make it faster?

For portrait mode: 1125x2048 window size, 1536x2436 frame size, fit inside and adjust window size option.

This gives you the following resolutions on common device sizes:
4:3 - 1536x2048
3:2 - 1365x2048
16:9 - 1152x2048
X - 1125x2436

This is a perfect fit for iOS devices such as ipad (1536x2048) and iphone x (1125x2436), but also a perfect fit for almost any Android device.

Little bit tricky, but it's possible to fetch the project folder and open it with android studio, then you can do any changes you want.
The easiest way to do it is edit the android version of any of the extensions you use and make an error on purpose, then when the compilation fails you can find it under C:\Users\*youruser*\AppData\Local\Temp
there will be a folder AND*something, just copy that out, and you can open it with Android Studio. Then just fix the error you made and change anything else you want, it takes some experience getting to know your way around android studio tho.

Oh, i think it happens when i load an animated .gif, i am now trying to filter out .gif files from being displayed, then i'll see in a day if it makes the crash go away from the 'last 24 hours' report on Google Play.

Sent you a PM back, still got issues, i also saw this in the logcat if it might help:

A/DEBUG: signal 11 (SIGSEGV), code 2 (SEGV_ACCERR), fault addr 0x74832000
A/DEBUG: backtrace:
A/DEBUG: #00 pc 000000000001ac60 /system/lib64/libc.so (memcpy+344)
A/DEBUG: #01 pc 000000000001055c /data/app/clickteam.softwarewolf.package-1/lib/arm64/libES3Renderer.so (_ZN6CImage13imageFillDataEPvii+112)
A/DEBUG: #02 pc 0000000000011fa8 /data/app/clickteam.softwarewolf.package-1/lib/arm64/libES3Renderer.so (Java_Banks_CImage_allocNative6+648)
A/DEBUG: #03 pc 00000000009ffd88 /data/app/clickteam.softwarewolf.package-1/oat/arm64/base.odex (offset 0x97b000)

Unfortunately i don't have a device where this crash can be reproduced, but i did get some strange artifacts when loading some images from URL into the active picture object on an Android 7.1.2 device.

Google Play reports these devices crashing: Please login to see this attachment. "signal 11 (SIGSEGV), code 2 (SEGV_ACCERR)"

If you have a new beta update that could resolve this crash i would be very happy if you could send it to me, i can see if i still get the graphical artifacts on my 7.1 device, and i can possibly also email a new .apk update to some users affected by the crash for them to try.
I always compile through Android Studio manually, so you could also give me the .java files/directories that should be replaced if it's easier.

Is anybody else seeing the "signal 11 (SIGSEGV), code 2 (SEGV_ACCERR)" in "memcpy" error?
A few versions ago didn't have this issue. I think maybe 292.17 or something like that was fine.
I am using a bunch of custom extensions, but i am not using CImage etc, so i think it's related to a standard extension, maybe the active picture object or something.

#00 pc 00000000000174bc /system/lib/libc.so (memcpy+100)
#01 pc 00000000000130d3 .../lib/arm/libES3Renderer.so (CImage::imageFillData(void*, int, int)+62)
#02 pc 0000000000013f65 .../lib/arm/libES3Renderer.so (Java_Banks_CImage_allocNative5+1260)

This only seems to happen on Android 7.0 and 7.1

I am also seeing a different crash which is only on Y5 2017 (HWMYA-L6737) Android 6.0: (common enough that it impacted 12 users tho)
#00 pc 000000000006a868 /system/lib64/libc.so (tgkill+8)
#01 pc 00000000000680b8 /system/lib64/libc.so (pthread_kill+68)
#02 pc 0000000000023a38 /system/lib64/libc.so (raise+28)
#03 pc 000000000001e1d8 /system/lib64/libc.so (abort+60)
#04 pc 0000000000433298 /system/lib64/libart.so (art::Runtime::Abort()+324)
#05 pc 0000000000138004 /system/lib64/libart.so (art::LogMessage::~LogMessage()+3136)
#06 pc 000000000030fe64 /system/lib64/libart.so (art::JavaVMExt::JniAbort(char const*, char const*)+2284)
#07 pc 0000000000310430 /system/lib64/libart.so (art::JavaVMExt::JniAbortF(char const*, char const*, ...)+224)
#08 pc 000000000034e4ac /system/lib64/libart.so (art::JNI::GetIntArrayElements(_JNIEnv*, _jintArray*, unsigned char*)+772)
#09 pc 00000000000f72f4 /system/lib64/libandroid_runtime.so (doDecode(_JNIEnv*, SkStreamRewindable*, _jobject*, _jobject*)+700)
#10 pc 00000000000f7d50 /system/lib64/libandroid_runtime.so
#11 pc 000000000386af00 /system/framework/arm64/boot.oat (offset 0x2941000)

I am getting new errors on Android from users after building with the latest 292.22 update, these did not occur with 292.17 which i believe my last release version was.
It's hard to know what is going on here, plausibly related to active image object?
It worked great on my device when i tested it, but some users are reporting crashes and freezes with latest app update.

Please login to see this attachment.