Changes

Bug 741 HiDPI: Fix OSX NEWT Offscreen Size Regression from commit 56d60b36798fa8dae48bf2aa5e2de6f3178ab0d1

Fix regression of commit 56d60b36798fa8dae48bf2aa5e2de6f3178ab0d1:
  createWindow(..) was issuing sizeChanged(..) to ensure size notification,
  however - the offscreen case used the dummy size 64x64.

  Fix issues the notifications in caller w/ true size.

Bug 741 HiDPI: Add missing window -> pixel unit conversion in AWTNewtEventFactory (e.g. for NewtCanvasAWT)

GraphUI: Fix label layout (absolute), add auto no-aa dpi threshold (200 dpi), add all modes to applet test-page

Bug 741 HiDPI: Fix missing window -> pixel unit conversion in AWTNewtEventFactory of commit 8b255eb303bba045b4eb087da1d1cb33b2e89e96

NEWT AWTAdapter: Simplify event processing / Window.enqueueEvent(..): Add API doc

Simplify event processing of NEWT's AWTAdapter:
  - enqueueEvent -> processEvent
  - add return value (new enum) determinating action

Using processEvent(..) now reduces manual invocation of NEWT Window.enqeueEvent(..).

Bug 741 HiDPI: Add new NativeSurfaceHolder interface to GLDrawable and NativeWindow; [AWT|SWT]NewtEventFactory use NativeSurfaceHolder as source, fixes pixel unit conversion

- Add new NativeSurfaceHolder interface to GLDrawable and NativeWindow, allowing NativeSurface access (pixel unit conversion)
  A NativeSurfaceHolder is e.g.:
    - NativeWindow (is-a)
      - NEWT [GL]Window

    - GLDrawable (has-a)
      - [AWT|SWT]GLCanvas

- [AWT|SWT]NewtEventFactory use NativeSurfaceHolder as source, fixes pixel unit conversion

UpstreamSurfaceHook: Add 'NativeSurface getUpstreamSurface()' (from EGLUpstreamSurfaceHook) to generalize ProxySurfaceImpl.getUpstreamSurface()

Minor: API-Doc / DEBUG Typo, GDISurface: Make fields private

Bug 741 HiDPI: Add ScalableSurface interface to get/set pixelScale w/ full OSX impl.

Add ScalableSurface interface
  - To set pixelScale before and after realization
  - To get pixelScale

  - Implemented on:
    - NEWT Window
      - Generic impl. in WindowImpl
      - OSX WindowDriver impl.
        - Also propagetes pixelScale to parent JAWTWindow if offscreen (NewtCanvasAWT)
      - AWT WindowDriver impl.

    - JAWTWindow / OSXCalayer
      - AWT GLCanvas
      - AWT GLJPanel
      - NEWTCanvasAWT:
        - Propagates NEWT Window's pixelScale to underlying JAWTWindow

    - WrappedSurface for pixelScale propagation
      using offscreen drawables, i.e. GLJPanel

    - Generic helper in SurfaceScaleUtils (nativewindow package)

  - Fully implemented on OSX

  - Capable to switch pixelScale before realization,
    i.e. native-creation, as well as on-the-fly.

  - Impl. uses int[2] for pixelScale to support
    non-uniform scale.

Test cases:
  - com.jogamp.opengl.test.junit.jogl.demos.es2.newt.TestGearsES2NEWT
  - com.jogamp.opengl.test.junit.jogl.demos.es2.awt.TestGearsES2AWT
  - com.jogamp.opengl.test.junit.jogl.demos.es2.awt.TestGearsES2GLJPanelAWT
  - com.jogamp.opengl.test.junit.jogl.demos.es2.newt.TestGearsES2NewtCanvasAWT

  - Press 'x' to toggle HiDPI
  - Commandline '-pixelScale <value>'
  - Added basic auto unit test (setting pre-realization)

Bug 741 HiDPI: Simplify ScalableSurface [set|get]SurfaceScale(..) spec, which also fixed JAWTWindow getSurfaceScale() issue on Windows

Let setSurfaceScale(..) return the validated requested values
and getSurfaceScale(..) always the current values.

This removes complication and solves a bug w/ JAWTWindow on Windows,
where we used 'drawable' as an indicator for 'previous locked' state.
The latter is not true since on Windows 'drawable' is set to null in unlock,
getWindowHandle() should be taken instead.

Bug 741 HiDPI: Simplify ScalableSurface (2): Add request pixelScale API entry, fixed NewtCanvasAWT use-case

We require the requested pixelScale in NewtCanvasAWT if the NEWT window (child)
is not yet realized, so the JAWTWindow can receive the request,
since realized/current pixelScale is still 1.

Remove return value (requested pixel scale):
  -  public int[] setSurfaceScale(final int[] result, final int[] pixelScale);
  +  public void setSurfaceScale(final int[] pixelScale);

Add API hook to query requested pixel scale:
  +  int[] getRequestedSurfaceScale(final int[] result);

Unique name for get[Current]*:
  -  public int[] getSurfaceScale(final int[] result);
  +  public int[] getCurrentSurfaceScale(final int[] result);

NEWT Window: Remove deprecated methods: HiDPI + reparentWindow(..)

Remove HiDPI pixel- from/to window-unit conversion and getter methods:
  Rectangle HiDPI pixel- from/to window-units are erroneous in case of multiple monitor setup where a mixed pixel-scale exist,
  since the methods didn't take the monitor viewport and each of it's pixel-scale into account (expensive).

Remove deprecated reparentWindow(..) methods.

Bug 741 HiDPI: Add ScalableSurface.getNativeSurfaceScale(..) to compute surface DPI ; Add NEWT Window.getPixelsPerMM(..) to query surface DPI

With HiDPI and surface scale, we need knowledge of the native surface's pixel-scale
matching the monitor's pixel-per-millimeter value.

Preserving the queried native pixel-scale and exposing it via
ScalableSurface.getNativeSurfaceScale(..) to compute surface DPI.

Add NEWT Window.getPixelsPerMM(..) to query surface DPI.

Surface DPI is demonstrated in GraphUI's GPUUISceneGLListener0A .. and TestRulerNEWT01, etc ..

GraphUI: Fix using multiple texture units w/ opt. colorTexUnit in GLRegion

The texture unit has to be updated always,
since program maybe used by multiple regions and diff. texUnits

GLMediaPlayer: Fix video stutter if using NullAudioSink

NullAudioSink shall return the last enqueued PTS in getPTS()
not causing a-v delta measure based on lagging audio in player.

Remove explicit xcode.clang property files

Bug 1011 / Bug 1012: GLMediaPlayer Audio/Video stuttering w/ OSX and OpenAL/JOAL (works using openal-soft default on all platforms now)

Graph: Fix GLSL issue w/ ES3, add 'const float sample_count', since define SAMPLE_COUNT is of type 'int'

Graph RegionRenderer: Fix GL3 and ES3 GLSL issues

Macro redefine 'texture2D -> texture'
was added _after_ the custom texture lookup insertion
causing GL3-core to fail.

GraphUI Demo: Analyze issue w/ OSX and NewtCanvasAWT when rendering / animating GLEventListenerButton

Analyze issue w/ OSX and NewtCanvasAWT when rendering / animating GLEventListenerButton.

FBO of main 'window' is _not_ updated ..

Probably related to CALayer - FBO - FBO* (of this button) ..

Fix Bug 1019 - Remedy of Bug 691 causes 'access/modify after free' and crashes the app

The 'magic' MyNSOpenGLContext::dealloc (MacOSXWindowSystemInterface-calayer.m)
of force destroying the underlying CGLContextObj of it's associated
NSOpenGLContext as introduced as a remedy of Bug 691 is plain wrong.

It was added in commit f6e6fab2a7ddfb5c9b614cb072c27ff697629161
to mitigate the experience behavior of delayed GL context
destruction when creating/destroying them multiple times
as exposed in unit test TestGLCanvasAddRemove01SwingAWT.

While this 'hack' worked for some reason on some OSX versions,
it caused a 'access/modify after free' issue exposed under some circumstances
and crashes the application.

The actual culprit of the delayed GL context destruction is different.
The offthread CALayer detachment and hence final destruction
issued on the main-thread is _not_ issued immediately
due to some referencing holding by NSApp.
Issuing an empty event on the NSApp (thread) will wake up the thread
and release claimed resources.

This has been found while realizing that the GL context
are released if the mouse is being moved (duh!).
This issue is also known when triggering stop on the NSApp (NEWT MainThread),
same remedy has been implemented here for a long time.

Fix NEWT EDTUtil Deadlock on EDTUtil.start()

DisplayImpl.runOnEDTIfAvail(..) issues EDTUtil.start()
while holding it's object-lock - if the EDT is not running,
then invokes the given task.

EDTUtil.start() impl. holds it's own edt-lock
while starting, then releases it's edt-lock while issuing a null-task.

If another thread injects a blocking task right in-between
which also acquires the display's object-lock it deadlocks.

Simply remove issuing the null-task, so EDTUtil.start()
can return immediatly (releasing edt-lock)
and allowing DisplayImpl.runOnEDTIfAvail(..)
also to release it's object-lock.

The other threads task then can be executed,
where the 'starting task' would come second - which is OK,
even though a rare occasion.

Above situation was triggered via AWT/NEWT reparenting w/ forced recreation
via TestParenting01dAWT.

+++

The null-task at EDTUtil.start() was remaining code to ensure
that the EDT completed starting, which is redundant.

Fix Bug 826 Regression caused by commit 41190c3830157abdf9649cbf7767e57108f55075 (Bug 975)

Commit 41190c3830157abdf9649cbf7767e57108f55075,
fix for 'Bug 975 GLJPanel's OffscreenDrawable double swap',
caused a regression of commit c427ed22244df44b71a0f1f000b0f93e56c283c2,
fix for 'Bug 826: GLJPanel: Fully restore TextureState and Viewport'.

Commit 41190c3830157abdf9649cbf7767e57108f55075 issues offscreenDrawable.swapBuffers()
and hence modifying the texture unit settings before saving the TextureState,
the whole purpose of commit c427ed22244df44b71a0f1f000b0f93e56c283c2.

Adapt to GlueGen commit c3054a01990e55ab35756ea23ab7d7c05f24dd37 (compount call-by-value extension), utilizing native JVMUtil_NewDirectByteBufferCopy(..)

Fix regression due to GlueGen commit c3054a01990e55ab35756ea23ab7d7c05f24dd37, completes JOGL commit 68ca3ae8fcce28c62034299bc6e6f7eaab50bd1f

Add handling of 'initializeImpl()' of generated GLU* classes:
  - GLU      : no 'initializeImpl()' <- no libs
  - GLUgl2es1: no 'initializeImpl()' <- no libs
  - GLUgl:     call 'initializeImpl()' in manual static init impl.

Further more, we shall not initialize the GLU* classes in GLU
static init method, querying GLU* class availability.

Tested via OlympicES1.

Adapt to GlueGen commit 2f6586292cd298bbc19d8acda0f7cf303c82078b (compount call-by-value extension)

- GLU*: initializeImpl() is no more required, since it's not forced .. uncomment in manual static initializer
- GLX: 'ForceStaticInitCode GLX', i.e. forcing native code, as used in our manual implementation.

Minor refinements: .gitignore HowToBuild

Bug 1021: Adding submodule oculusvr-sdk and adjust HowToBuild (git clone --recurse-submodules <jogl-url>)

Bug 1021: Add OculusSDK binding / Basic OVR support

Bring-up test only initializes the ovrHmdHandle
data structure by OVR native code.
    See com.jogamp.oculusvr.OVRVersion

Current evaluation build compiles all OVR-SDK source
files itself w/o using provided libovr.a.
We also skip the GL dependent renderer of the SDK,
i.e. we prefer to utilize our JOGL 'barrel distortion' renderer.
This eases the 'chicken-egg' problem of OVR SDK dependencies,
i.e. libGL, libGLESv2 .. etc

Since the OVR source code is C++, we may still have to figure out
how to / and whether we shall link libstdc++ statically
to remove platform dependencies.
Right now we link libstdc++ statically if using GCC,
see  make/build-oculusvr.xml (hackish .. TODO: better way to include all symbols).

Same consideration applies to GNU/Linux and libudev.so dependency,
since there are:
    - libudev.so.0 and (Older distri's)
    - libudev.so.1 (Debian8, ..)

...

Produced JAR artifacts are
    - jar/atomic/oculusvr.jar
    - jar/atomic/oculusvr-natives-<os.and.arch>.jar

i.e. only in 'atomic' variants to not bloat the default 'all' JAR files.

....

make/build-oculusvr.xml Notes:
    - Currently native build only enabled on GNU/Linux (isLinux)
    - Force disable native build via property 'c.build.oculusvr.skip'

Bug 1021: Add OculusSDK binding / Basic OVR support (Refine build exclusion)

common:
  - property 'oculusvr.sdk.available' is only set if sub-module 'oculusvr-sdk' is available
  - if 'oculusvr.sdk.available' is not set build.xml will not issue build-oculusvr.xml's 'all' target

build-oculusvr.xml:
  - propery 'oculusvr.build.enabled' is only enabled if 'oculusvr.sdk.available' is set
    and targer 'all' will only pass if set.

Bug 1021: Add OculusSDK binding / Basic OVR support (HowtoBuild.xml: Mention feature loss if not fetching submodules)

Bug 1021: Add OculusSDK binding / Basic OVR support (Bring-up on Windows w/ Mingw64 ; Solve static linkage of stdc++)

- Bring-up on Windows w/ Mingw64
  - Bumped oculusvr-sdk to commit 70e44b846ceaf9eebc336b12a07406c3a418a5c3
  - Added required mingw64 linker statments
  - Builds and runs on x86_64 and i386

- Solve static linkage of stdc++
  - "-Wl,-Bstatic -lstdc++ -Wl,-Bdynamic" MUST come AFTER all object files!!
  - No more manual '--undefined=symbol' required .. puhh :)

- Cleanup linker cfg ..

- Still runs on GNU/Linux x86_64 and i386

Bug 1021: Add OculusSDK binding (Bring-up on OSX w/ clang/clang++ ; Split native build: cc, c++, linkage)

- Bring-up on OSX w/ clang/clang++
  - Build and tested on OSX 10.9

- Split native build: cc, c++, linkage
  - Perform steps sequentually to not
    differentiate c and c++ compiler as required for clang/clang++
    - compile c code
    - compile c++ code
    - link all object files

Bug 1021: Add note about 'Oculus VR Rift SDK Software License' in LICENSE.txt

WWW: Update CCT's C3D Entry

www: shorten c3d desc.

www .. c3d desc (2)

Bug 1021: Enhance OVR GlueGen Mapping / Adapt to GlueGen commit 9ee44e1a289ecbac024662dd5a2ffc42e8add023 (Bug 1025)

Bug 1021: OVR GlueGen Mapping: Compile all generated C files, add missing C ovrHmdStruct in generated files, reduce warnings.

Bug 1021: OVR GlueGen Mapping: Handle non-existent native oculusvr lib gracefully

- query isAvailable() in ovr_Initialize(), ovrHmd_Create(..) and ovrHmd_CreateDebug(..)
  and return appropriate values.

Fix build if oculusvr-sdk submodule is missing

GLU: Make ProjectFloat/ProjectDouble final and deprecate GLU.destroy() method.

Enhance FloatUtil: Merge ProjectFloat features while adding optimized variations; PMVMatrix: Remove NIO buffer path, use backing-array only and FloatUtil direct.

- FloatUtil pptimized variants:
  - mapObjToWinCoords (gluProject) passing 'P x Mv' for batch operations

  - mapWinToObjCoords (gluUnProject) passing 'Inv(P x Mv)' for batch operations

  - mapWinToObjCoords (gluUnProject) passing 'Inv(P x Mv)' and two winz values
    for two ray picking resulting in two obj positions. (-> mapWinToRay)

- PMVMatrix
  - dropped low performance NIO mode
  - simply use common backing-array and fixed offsets directly
  - drop ProjectFloat usage in favor of FloatUtil
  - reduce number of temporary arrays

Enhance FloatUtil: More optimizations, concludes commit 0bded476868c5fdfe44502bfd55957469d0d72bb

FloatUtil optimizations (unroll and linear memeory access):
  - transposeMatrix
  - invertMatrix (diff algo as well - 50% speed bump)
  - multMatrix
  - multMatrixVec

FloatUtil added
  - matrixDeterminant(..)

FloatUtil removed
  - Certain FloatBuffer variants are removed
    or at least marked deprecated.

Add missing FloatUtil.invertMatrix(..) unit test, which impl. changed w/ commit ee774dce9e474e8ea961bd9b504d26e9321e1b15

WIP: Add Matrix4 OO wraper for FloatUtil matrix operations

Fix 'typo' in messages: 'Catched' -> 'Caught'

OculusVR: Add in eclipse classpath, and jogl-java-src.zip; Also create jogl-test-java-src.zip

Bug 1027: Fix usage of atomic jar files (NEWT)

Issue was that 'com.jogamp.opengl.util.PNGPixelRect'
was included in jogl-core.jar and that NEWT assumes
PNG conversion is possible having this class available.

However, PNGPixelRect requires 'jogamp.opengl.util.pngj'
which is only included within jogl-util.jar.

Moved PNGPixelRect from jogl-core.jar to jogl-util.jar.

+++

Added manual test launch 'testnoawtatomics' in make/scripts/tests.sh
to test atomic usage.
Works now w/ com.jogamp.opengl.test.junit.jogl.demos.es2.newt.TestGearsES2NEWT,
showing that the window/application icon is _not_ set due to not having PNG* available.

Fix (c) header: Due to copy & paste, wrong header was used for certain files .. (sorry)

Bug 1021: Add OculusVR distortion renderer (single FBO and dual FBO); Add GLEventListener2 (WIP); Refine FloatUtil

- GLEventListener2 extends GLEventListener adds refined control:
  - display w/ flags, i.e. repeat, don't clear
  - setProjectionModelview(..)

- FloatUtil.* Add return value for chaining, where missing

+++

- jogamp.opengl.oculusvr.OVRDistortion
  - Handles all OVR related data and maps it to shader + GL buffers
  - display method

- com.jogamp.opengl.oculusvr.OVRSBSRendererSingleFBO implements GLEventListener
  - Simple OVRDistortion renderer using single FBO
  - Using upstream GLEventListener2 (the content)

- com.jogamp.opengl.oculusvr.OVRSBSRendererDualFBO implements GLEventListener
  - Simple OVRDistortion renderer using two FBOs
  - Using upstream GLEventListener2 (the content)

Manual Test: com.jogamp.opengl.test.junit.jogl.stereo.ovr.OVRDemo01

Bug 1021: Refine Stereo Rendering API and OculusVR implementing renderer

Refine API in regards to proper package names, interface
and high-level access to eye specific constant parameter
and variable eye movement.

+++

Commit 36327e24cf586b50bf18e87d7d13d53eb41cf1d9 introduced 'GLEventListener2'

Move javax.media.opengl.GLEventListener2
  -> com.jogamp.opengl.util.CustomRendererListener
    -> com.jogamp.opengl.util.stereo.StereoRendererListener

StereoRendererListener adds stereoscopic specific:

  public void reshapeEye(final GLAutoDrawable drawable, final int x, final int y, final int width, final int height,
                         final EyeParameter eyeParam, final EyePose eyePose);

.. see below.

++

Add com.jogamp.opengl.util.stereo:
  - EyeParameter (Constant eye parameters, like IPD and FOV)
  - EyePose (Current eye position and orientation)

+++

Add com.jogamp.opengl.math.FovHVHalves to support
non-centered bi-directional FOV for lenses.

Add respective FloatUtil.makePerspective(.. FovHVHalves fovhv ) variant.

+++

Bug 1021: OVRDistortion.getSBSUpstreamPMV(..): Use generic perspective matrix constructor (example implementation only)

Bug 1021: OVRDistortion.getSBSUpstreamPMV(..): Use all cached eye parameter (example implementation only)

Bug 1021: OVRSBSRenderer*: Add texture filtering parameter (bilinear/none), adjust FBO texture params (alpha, filtering); Demo: Pass all parameters w/ arguments, use sane good defaults.

Turns out that using bilinear texture filtering w/o multisampling is 'good enough' ..

- OVRSBSRenderer*:
  - Add texture filtering parameter (bilinear/none)

  - Adjust FBO texture params (alpha, filtering)
    No alpha needed in FBO sink/target!

- Demo: Pass all parameters w/ arguments, use sane good defaults.

FloatUtil.makePick(..): Refine API doc, incl. PMVMatrix.gluPickMatrix(..)

Code Clean-Up based on our Recommended Settings (jogamp-scripting c47bc86ae2ee268a1f38c5580d11f93d7f8d6e74)

- Change non static accesses to static members using declaring type
        - Change indirect accesses to static members to direct accesses (accesses through subtypes)
        - Add final modifier to private fields
        - Add final modifier to method parameters
        - Add final modifier to local variables
        - Remove unnecessary casts
        - Remove unnecessary '$NON-NLS$' tags
        - Remove trailing white spaces on all lines