of joysticks and gamepads. It now presents a gamepad
on the screen that updates itself whenever any connected
joystick or gamepad are used. This allows the tester
to compare the actual controller layout to the "default"
controller layout.
Information about the active joystick's available axes
and buttons is also rendered to the display.
A dump of all controllers and their components is written
to joysticks-###.txt file.
git-svn-id: https://jmonkeyengine.googlecode.com/svn/trunk@9762 75d07b2b-3a1a-0410-a2c5-0572b91ccdca
- re introduced the alphaDiscardThreshold as explained in prvious commit. It's not binded to the AlphaTestFallOff fixedfunc binding
- Added a small poly offset to post shadow technique, this greatly help in fixing shadow acne.
- Added Poisson disc sampling PCF Filtering for shadows
- Properly passed the shadow map size as a define in the shaders and remove the hardcoded value
- Pssm15 don't use the textureSize function anymore and use the same shadow map size define ( this increased performance quite a bit)
- Optimized the shaders code a bit
- Better PSSM test
git-svn-id: https://jmonkeyengine.googlecode.com/svn/trunk@9750 75d07b2b-3a1a-0410-a2c5-0572b91ccdca
* TestImageRaster tests both reading and writing from various types of formats now by chaining the image conversions
*
git-svn-id: https://jmonkeyengine.googlecode.com/svn/trunk@9710 75d07b2b-3a1a-0410-a2c5-0572b91ccdca
Old XXXTrack classes are now deprecated an just extend the corresponding XXXEvent class.
Also removed PositionTrack, RotationTrack and ScaleTrack for they are deprecated since alpha 3.
git-svn-id: https://jmonkeyengine.googlecode.com/svn/trunk@9567 75d07b2b-3a1a-0410-a2c5-0572b91ccdca
* Now uses only OpenGL2 draw buffers instead of relying on GL_ARB_draw_buffers existing
* copyFrameBuffer() was non-functional when used against the main framebuffer because the width/height were set to zero erroneously, now it uses the current viewport parameters.
* Added TestRenderToCubemap to demonstrate render to cubemap functionality, it also uses MRT to render to all the cube sides
git-svn-id: https://jmonkeyengine.googlecode.com/svn/trunk@9378 75d07b2b-3a1a-0410-a2c5-0572b91ccdca
* DesktopAssetManager now works as before (flipY and generate mips options are enabled)
* Fix crash in TestSimpleWater
git-svn-id: https://jmonkeyengine.googlecode.com/svn/trunk@9322 75d07b2b-3a1a-0410-a2c5-0572b91ccdca
* Added 3 implementations of AssetCache: SimpleAssetCache, WeakRefAssetCache and WeakRefCloneAssetCache
* Added AssetProcessor interface that handles cloning and post processing of assets after they are loaded
* AssetKey can now configure which cache/processor to use for a particular asset type
* Added AssetManager unregisterLoader method
* AssetManager now supports more than one AssetLoadListener
* Javadoc improvements in AssetManager
* Asset interface now renamed to CloneableSmartAsset (which more accurately describes its behavior and use case)
* DesktopAssetManager now makes proper use of the new AssetProcessor/AssetCache classes when handling asset loading
* Added proper equals/hashCode methods to many AssetKey subclasses, which is required for the new system to work properly
* All AssetKeys were rewritten to work with the new asset system
* loadAsset(AudioKey) now returns an AudioNode and not AudioData, similar to the behavior of loadAsset(TextureKey) returning a Texture and not an Image. Because of that, the key storage in AudioData has been removed.
* Texture, Spatial, and Material are all cloneable smart assets now and will be cleared from the cache when all instances become unreachable
* Improved the existing TestAssetCache test to make sure the new system works
git-svn-id: https://jmonkeyengine.googlecode.com/svn/trunk@9309 75d07b2b-3a1a-0410-a2c5-0572b91ccdca
This allow to have shadows that fully works with partially transparent objects (like trees).
If a material does not have the postShadow technique, the renderer uses the fallback material (like before).
git-svn-id: https://jmonkeyengine.googlecode.com/svn/trunk@9279 75d07b2b-3a1a-0410-a2c5-0572b91ccdca
and replaced it with camNode.setEnabled().
A) it is kind of scary to rely on the control index in
this case.
B) CameraNode.setEnabled() already calls setEnabled()
on the CameraControl.
An acceptable alternative would be:
camNode.getControl(CameraControl.class).setEnabled()
but I didn't see the point of it.
git-svn-id: https://jmonkeyengine.googlecode.com/svn/trunk@9267 75d07b2b-3a1a-0410-a2c5-0572b91ccdca
Sha..om
PSp..om
rem..om
bre..om
pot..om
roo..li
dan..om
nor..67