[java] viewPortChanged 1280x960 / 2.0 -> 640.0x480.0
     [java] Exception in thread "main" java.lang.NoSuchMethodError: java.nio.IntBuffer.position(I)Ljava/nio/IntBuffer;
     [java] 	at playn.core.GL20$Buffers.resizeIntBuffer(GL20.java:397)
     [java] 	at playn.core.GL20.glGetShaderiv(GL20.java:631)
     [java] 	at playn.core.GLProgram.compileShader(GLProgram.java:112)
     [java] 	at playn.core.GLProgram.<init>(GLProgram.java:45)
     [java] 	at playn.core.TriangleBatch.<init>(TriangleBatch.java:122)
     [java] 	at playn.core.TriangleBatch.<init>(TriangleBatch.java:114)
     [java] 	at playn.scene.SceneGame.createDefaultBatch(SceneGame.java:95)
     [java] 	at playn.scene.SceneGame.<init>(SceneGame.java:38)
     [java] 	at rook.core.MainGame.<init>(MainGame.java:25)
     [java] 	at rook.java.JavaMain.main(JavaMain.java:13)
     [java] Java Result: 1

This seems to be an issue with lwjgl which is fixed since 3.1.5. I locally updated lwjgl in playn to 3.1.6 and the error disappears.

UPDATE just recompiling playn locally also makes the error disappear, so it seems to have nothing to do with lwjgl. But rather with playn being built with java 9 (as far as I can tell from the MANIFEST) and me using java 8 locally.

When manually setting scaleFactor to 1 on a hidpi device (MacBook Pro Retina 2015), mouse input coordinates become broken. I found the reason for this in HtmlGraphics:

mouseScale = config.scaleFactor / HtmlPlatform.devicePixelRatio();

disabling this workaround (and setting mouseScale to 1) actually fixes my issue. However my game is using a custom scaling method to always fully occupy the available window/space so I'm not sure if that intereferes with this workaround.

If I don't change scaleFactor and leave it on its local default value of 2, then I have no problem with mouse coordinates.

Based on this it seems to me that this mouseScale workaround is unnecessary.

Tested on Google Chrome and Safari.

Regions were caching tile instances which may refer to textures that may already have been disposed. Now this is detected when tile() is called and a new tile is created from the new texture.

Hi, I'm tying to configure gradle for playn project, instead of maven. I found that gradle can't resolve "apk-extention" dependency:

FAILURE: Build failed with an exception.

* What went wrong:
A problem occurred configuring project ':android'.
> Could not resolve all dependencies for configuration ':android:_debugApkCopy'.
   > Could not find com.google.android:apk-expansion:r3.
     Required by:
         project :android > io.playn:playn-android:2.0-rc3

If checkmaven central for this package I see 404: [404] Not Found: /artifact/com.google.android/apk-expansion

Workaround for gradle based android builds is to exclude this dependency. It works fine. But maven users must see an error.

While letting the browser control frame render times is a nicer solution, setting a lower fps manually results in your CPU not overheating if rendering is slower than 60 fps, which is ultimately pretty useful.

• some cleanup of dead code

I noticed that getText had hardcoded "Ok" and "Cancel" as labels for the buttons (like sysDialog does). This PR fixes it, while being backwards compatible with those labels as defaults

Google recently announced that future updates of the applications must include 64-bit version of native libraries. To be honest, I don't know what libplayn-android-nativelib.so does, maybe it's something very old, but it still part of the PlayN android module.

I am using StartApp ads with my game. After showing a rewarded video some parts of game textures become black. I think it might be opengl context problem. how to solve it? here is my game : https://github.com/beshkoon/beshkoon/raw/master/beshkoon.github.io/amintest-release.apk

playn.core.Input has a set of boolean flags (mouseEnabled, touchEnabled, keyboardEnabled) which are supposed to enable or disable dispatching of the corresponding types of events. tripleplay.game.ScreenStack actually uses this to disable dispatching of input events during transitions.

However this does not seem to work; nobody seems to be checking the state of these variables before dispatching events.

Indeed, while testing on the Java and Android backends, I see that events are dispatched during screen transitions, even though they should not be.

Using any browser on a touch screen Microsoft laptop, mouse events are not sent to PlayN when the input device is the trackpad or a mouse. When the input device is the touchscreen, however, events are sent and received as expected.

I have observed this issue on a few different Lenovo Yoga models and believe it extends to all PCs with touch screens based on our customer bug reports.

Hi, All. I'm new using PlayN. 1 - I have generated a sample project with the following command (mvn archetype:generate -DarchetypeGroupId=io.playn -DarchetypeArtifactId=playn-archetype -DarchetypeVersion=2.0) The problem is that when I try to compile the project from console (mvn test -e -X) I get the following error:

error-maven-android.txt

Do you have any idea how I can solve it?

2 - I've also tried compiling the playn-samples in eclipse (Neon) but it also generates an error

Plugin execution not covered by lifecycle configuration: com.simpligility.maven.plugins:android-maven-plugin:4.2.1:emma (execution: default-emma, phase: process-classes) pom.xml /playn-hello-android

Versions: maven-3.5.0 Java version: 1.8 (Oracle)

[Edit update error]

In some configurations, sound clips can only be played once.

• My environment: OpenJDK 1.8.0_141 on Ubuntu 14.04.5.
• IcedTea/PulseAudio sound provider (default for OpenJDK in Ubuntu)
• PlayN 2.1-SNAPSHOT

Sample code:

public class SoundTest extends SceneGame {
  boolean playAgain = true;
  playn.core.Sound sound;

  public SoundTest (final Platform plat) {
    super(plat, 33); // update our "simulation" 33ms (30 times per second)

    // create and add background image layer
    Image bgImage = plat.assets().getImage("images/bg.png");
    ImageLayer bgLayer = new ImageLayer(bgImage);
    // scale the background to fill the screen
    bgLayer.setSize(plat.graphics().viewSize);
    rootLayer.add(bgLayer);
    
    sound = plat.assets().getSound("sounds/drop");
    sound.play();
  }
  
  public void update(Clock t) {
    if (playAgain && plat.tick() > 5000) {
      sound.play();
      playAgain = false;
    }
  }
}

This may be related to the following:

• https://bugs.openjdk.java.net/browse/JDK-8077019
• https://ubuntuforums.org/showthread.php?t=1903727

Added profile -Pall for all artifacts (except the test-bugvm that might work only from osx). Added -DskipExec to be able to deploy without executing any test main apps.

Note sure what is causing this issue, but thought I'd share that the SDM compiles aren't working. But a regular compile I believe works fine.

Compile with -strict or with -logLevel set to TRACE or DEBUG to see all errors.
         Tracing compile failure path for type 'pythagoras.util.Platform'
            [ERROR] Errors in 'jar:file:/C:/Users/Ben/.m2/repository/com/samskivert/pythagoras/1.4.2/pythagoras-1.4.2-sources.jar!/pythagoras/util/Platform.java'
               [ERROR] Line 52: Cannot invoke clone() on the array type double[]
               [ERROR] Line 31: Cannot invoke clone() on the array type byte[]
               [ERROR] Line 45: Cannot invoke clone() on the array type float[]
               [ERROR] Line 38: Cannot invoke clone() on the array type int[]
         [ERROR] Hint: Your source appears not to live underneath a subpackage called 'client'; no problem, but you'll need to use the <source> directive in your module to make it accessible

Wait! Wtf?

Well, some javascript capability can help dev teams to accelerate dev of a complex game:

1. Give some power to artistis (animations)
2. Put logic on scripts can help non-developers to help.

Currently, java, android, html, and iOS have javascript engines (a bridge between "native" and js code).

The idea is not create a powerfull script engine on PlayN, but just support the use of javascript on all platforms.

What do you think?