@BINALECTRUM said in Android build and deploy. 64bit, Clang, new app store developer console rules [2018 & 2019]:

Is Qt going more in embedded/car side and just slowly drops Android/mobile platform? In Roadmap 2018 there is nothing new about mobile platform.

i heavily doubt that.
I guess the reason is more because Qt runs more or less stable on Android. There are no bigger features which need to be implemented, mostly bug fixing.

What is your move Qt team on new app store rules in developer console: minimal platform-26 in August 2018 and 64-bit apk in 2019? Some new changes in build and deploy area?

i asked this at the beginning of the year, but unfortunately haven't received an answer.

@noone Would this be of any help to you? https://forum.qt.io/topic/79730/no-build-step-for-android

It seems this guy had the same problem and solved it with adding gradle.

Usually that error has to do with missing ndk/sdk or the more likely ant/gradle. Try running the build or deploy on the command line and see if it gives you more information.

Then it's strange, OAuth is the third protocol in the list of the Apereo CAS Implementation part of the article you linked.

If found the way by myself, if this can help to someone, is a question of attaching the texture again and change it before the fbo generation.

Like:
```
glBindTexture(GL_TEXTURE_2D, d->fbo->texture());
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_LINEAR); // Linear Filtering
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, GL_LINEAR); // Linear Filtering

Hi,

AFAIK, no. Modifying system settings like that is something that is outside of Qt's scope (I don't meant you can't do it using Qt but it has no direct support for that).

Hi @tansgumus ,

Apple is very conservative in allowing developers long term background threats/events.

Taken from the Apple docu:

Implementing Long-Running Background Tasks For tasks that require more execution time to implement, you must request specific permissions to run them in the background without their being suspended. In iOS, only specific app types are allowed to run in the background: Apps that play audible content to the user while in the background, such as a music player app Apps that keep users informed of their location at all times, such as a navigation app Apps that support Voice over Internet Protocol (VoIP) Newsstand apps that need to download and process new content Apps that receive regular update from external accessories Apps that implement these services must declare the services they support and use system frameworks to implement the relevant aspects of those services. Declaring the services lets the system know which services you use, but in some cases it is the system frameworks that actually prevent your application from being suspended.

That said, a navigation app should fall in this category. But as it is, you're bound to do it in OBjective-C but its relative easy to include ObjectiveC code in Qt.

I solved it for now by using a size relative to its parent and calculating a scale from it. Also applying the same scale to the position of the Pathview gives a usable solution, but it requires a few additional calculations.
I see how the (617 * 411) resolution occurs.
So the Windows width: and height: properties are not sizes in pixels..

Desktop - 1920*1200 or 2560x1440
density 4.2969047380546925
pixelRatio 1

Mobile
density 6.5092731829573935 - 1620 x 1080
pixelRatio 2.626672502927989

1620/2.6266... gives 617
1080/2.6266... gives 411

Hi,

No it's not. Your application becomes a library loaded by Android's runtime which requires a small amount of Java but you are writing and using C++. You can access Android's Java functionalities through JNI if needed.

@Gourmet He is building for Android using the official compiler from NDK I guess, not MinGW.
And why do you think MinGW is "castrated"? It is simply GCC ported to Windows.

Okay, I've just seen in the Building for android page that at most NDK r10 must be used. I recompiled qt with it and targeting android-21 platform. It works with QMake but with CMake I still get the same error though. Quite a step ahead fortunately :)

AFAIK, no. But you can check that yourself.

Hello,

Sorry for the novice mistake! I was trying to use different compilers in the NDK but i end up with the same error. The code i uploaded before was from one of my trials.
I did change my configuration to x86 but i still get the same error. Since this is my 1st time for android, I may have missed something.

The error stays the same. The 1st few line of my cmake error log states:

Compiling the CXX compiler identification source file "CMakeCXXCompilerId.cpp" failed. Compiler: /home/user/android-ndk-r16b/toolchains/x86-4.9/prebuilt/linux-x86_64/bin/i686-linux-android-g++ Build flags: Id flags: The output was: 1 /home/user/android-ndk-r16b/toolchains/x86-4.9/prebuilt/linux-x86_64/bin/../lib/gcc/i686-linux-android/4.9.x/../../../../i686-linux-android/bin/ld: error: cannot open crtbegin_dynamic.o: No such file or directory /home/user/android-ndk-r16b/toolchains/x86-4.9/prebuilt/linux-x86_64/bin/../lib/gcc/i686-linux-android/4.9.x/../../../../i686-linux-android/bin/ld: error: cannot open crtend_android.o: No such file or directory /home/user/android-ndk-r16b/toolchains/x86-4.9/prebuilt/linux-x86_64/bin/../lib/gcc/i686-linux-android/4.9.x/../../../../i686-linux-android/bin/ld: error: cannot find -lstdc++ /home/user/android-ndk-r16b/toolchains/x86-4.9/prebuilt/linux-x86_64/bin/../lib/gcc/i686-linux-android/4.9.x/../../../../i686-linux-android/bin/ld: error: cannot find -lm /home/user/android-ndk-r16b/toolchains/x86-4.9/prebuilt/linux-x86_64/bin/../lib/gcc/i686-linux-android/4.9.x/../../../../i686-linux-android/bin/ld: error: cannot find -lc /home/user/android-ndk-r16b/toolchains/x86-4.9/prebuilt/linux-x86_64/bin/../lib/gcc/i686-linux-android/4.9.x/../../../../i686-linux-android/bin/ld: error: cannot find -ldl collect2: error: ld returned 1 exit status

and the cmake output states

-- The CXX compiler identification is GNU 4.9.0 -- Check for working CXX compiler: /home/user/android-ndk-r16b/toolchains/x86-4.9/prebuilt/linux-x86_64/bin/i686-linux-android-gcc -- Check for working CXX compiler: /home/user/android-ndk-r16b/toolchains/x86-4.9/prebuilt/linux-x86_64/bin/i686-linux-android-gcc -- broken CMake Error at /usr/share/cmake-3.5/Modules/CMakeTestCXXCompiler.cmake:54 (message): The C++ compiler "/home/user/android-ndk-r16b/toolchains/x86-4.9/prebuilt/linux-x86_64/bin/i686-linux-android-gcc" is not able to compile a simple test program. It fails with the following output: Change Dir: /home/user/testProjects/build-bubbleWork-Android_for_x86_GCC_4_9_Qt_5_9_3_for_Android_x86-Default/CMakeFiles/CMakeTmp Run Build Command:"/usr/bin/make" "cmTC_c05c5/fast" /usr/bin/make -f CMakeFiles/cmTC_c05c5.dir/build.make CMakeFiles/cmTC_c05c5.dir/build make[1]: Entering directory '/home/user/testProjects/build-bubbleWork-Android_for_x86_GCC_4_9_Qt_5_9_3_for_Android_x86-Default/CMakeFiles/CMakeTmp' Building CXX object CMakeFiles/cmTC_c05c5.dir/testCXXCompiler.cxx.o /home/user/android-ndk-r16b/toolchains/x86-4.9/prebuilt/linux-x86_64/bin/i686-linux-android-gcc -o CMakeFiles/cmTC_c05c5.dir/testCXXCompiler.cxx.o -c /home/user/testProjects/build-bubbleWork-Android_for_x86_GCC_4_9_Qt_5_9_3_for_Android_x86-Default/CMakeFiles/CMakeTmp/testCXXCompiler.cxx Linking CXX executable cmTC_c05c5 /usr/bin/cmake -E cmake_link_script CMakeFiles/cmTC_c05c5.dir/link.txt --verbose=1 /home/user/android-ndk-r16b/toolchains/x86-4.9/prebuilt/linux-x86_64/bin/i686-linux-android-gcc CMakeFiles/cmTC_c05c5.dir/testCXXCompiler.cxx.o -o cmTC_c05c5 -rdynamic /home/user/android-ndk-r16b/toolchains/x86-4.9/prebuilt/linux-x86_64/bin/../lib/gcc/i686-linux-android/4.9.x/../../../../i686-linux-android/bin/ld: error: cannot open crtbegin_dynamic.o: No such file or directory

Hi,
for an easy native-styled Android menu (QML-based though), you can also have a look at V-Play Engine, which extends Qt and handles native features and styling for Android and iOS.

E.g. for a native styled Navigation, with the V-Play Navigation Components:

import VPlayApps 1.0 App { Navigation { NavigationItem { title: "Home" icon: IconType.home NavigationStack { Page { title: "Home" } } } NavigationItem { title: "Favorites" icon: IconType.star NavigationStack { Page { title: "Favorites" } } } } }

Best,
GTDev

Dear Allerknapper
I have same problem and stacked long time at here .
Now I have use the QT5.10.1 maybe can get succeed.
I have no time on it, but i will try it from July.
If you have any new progress could you update on this topic?
Best regards,
jeff

@KroMignon said in Qt Android project and CVS like Git, Perforce, etc.:

Does it make sense to update it to a newer version (eg 3.5.1)?

Uh, no idae, to be honest. I've never installed Gradle myself. I think it comes bundled with Android SDK nowadays.

The backends are different between platforms so what you observe currently with Android is likely unique to Android.

For anyone finding this thread via Google OpenSSL actually has a somewhat detailed description about the workaround: https://wiki.openssl.org/index.php/Android#Wrapper_Shared_Objects

Hi,

You can fill a suggestion for updating the documentation on the bug report system.

Note that the Android folks have a tendency to make new releases that are adding/removing/moving stuff around and are breaking things here and there thus the documentation might not be accurate because of that.

@GrecKo I found that it is a bug that is affecting all platforms with certain screen resolutions. Some developers may not notice this issue but the end user may be using a certain screen resolution and be affected by the issue! You can login to https://bugreports.qt.io/ and Vote THIS bug if you also want this bug to be resolved.

Here is the solution.

http://doc.qt.io/qt-5/publishtogoogleplay.html

Thanks everyone for insight . I have some idea now from where i should start .

Tested with Qt 5.9.2 and it's working as expected. Well, the script bash is called.

@torpeanders I always use just one resource file per project, and although not completely sure, it looks like having RESOURCES += ...
is intended for just one resource file.
I'm not sure if you need two resource files, I'd ask what is your use case for that, please remember that you can have subfolders to split just in case; i.e. I usually have a subfolder images for the icons and a subfolder translations and then it's a question of using the proper path

QIcon(":/images/cut.png")

So if even this trick cannot avoid you having two resource files, please have a look at the documentation as it explains how to create/deal with external resource files (it'll be your 2nd one).

@JKSH That's it

Thank you guys.

Alright, problem solved. As so many times, the bug was sitting between chair and keyboard.

In the FontLoader, I had the name property set. Which is set automatically if a font is loaded by source. https://doc.qt.io/qt-5/qml-qtquick-fontloader.html

Having the name property set while loading a font by source is no problem under macOS but under Android, one Devanagari font started to dominate all instances, where Devanagari symbols appeared and all logic, how fonts are applied falls apart.

@Attila-Santo
i will release the plugin for testing purposes during alpha/beta phase at some point when it makes sense.
Watch this thread to get the notification.

Oh dear, you're absolutely right. I was completely overdoing it.
In that case, I must only check for the language files to be present.

@jweber

Hi and welcome to devnet forum

I am not sure if this is possible at all. It is quite a long time back when I used MSVC. However, I am having no good feelings when I read your post.

I had started recently compiling for Android. Since a couple of years I am using Qt creator and MinGW compiler. The Android compiling process is GCC-based as MinGW is. That is taking some ambiguities out. Therefore my recommendation you may save time and conserve some hair, when you are going directly to Qt creator when starting Qt for Android.

The error message you are presenting is probably because of the mix MSVC and GCC-make, which might errornously recognized as linux version. I might be wrong there, but this looks fishy.

In case you decide to follow my advice, also go directly to Qt 5.9.0 (at least). There seem to be recent change also on the Android side, which was not really helpful in getting started. I had struggled quite a bit when trying to follow the different advices and tutorials. "gradle" seems to have replaced "ant" already on Android. At the end I had decided to go directly to newest Qt5.9.0 at that time instead of fiddling around with the previous stuff. That went much easier.