Changes

Fix javadoc package property, use comma as separator (dropped newt and oculusvr apidoc)
 (commit: 00cb1f6)

GLStateTracker: Avoid private access wrapper, reuse fetched pixelStateMap
 (commit: 215cbe5)

Scripts: Bump to java 1.7.0_67 and apache-ant 1.9.4
 (commit: e3226d1)

Fix commit b5910f18f0b82a8a1f6f6252dc19971d5e487f39 for toString(): Call super.hashCode(), due to InternalError("hashCode not designed")
 (commit: c28b5eb)

graph/font: Use 'pixelSize' var-name, remove redundancies
 (commit: 93862d1)

Graph/Text: Prepare unit tests for font rendering validation (WIP)
 (commit: fc6df2f)

Graph/Text: Prepare unit tests for font rendering validation (WIP)
 (commit: 9c9d0cf)

HowToBuild: Bump requirements: ant >= 1.9.0, git >= 1.7.10

- ant >= 1.9.0
 - Due to java8 build support

- git >= 1.7.10
 - Due to gluegen's new git query of HEAD branch and sha1-tip
 (commit: 1882b65)

HowToBuild: Revert required git version back to 1.6.0 (from 1.7.10, commit 1882b6551326e583c311b6b169be2222125df4d9)

git version 1.7.10 is not available on certail 'long life' / LTS systems,
we have to reschedule the gluegen patch for a later time.
 (commit: 8ca6a47)

TestVersionSemantics: BACKWARD_COMPATIBLE_USER current vs 2.2.0
 (commit: 262592b)

glsl.sdk.CompileShader: Use IOUtil.StreamMonitor
 (commit: 28ad09d)

GLSharedContextSetter: Fix API doc and all html references (fix URL and complete notes in all implementations)
 (commit: 4dfa5e3)

Bug 1054: Cleanup GLContext 'lock' and 'drawable' usage, perform drawable null check in constructor.

This patch merely cleans up 'lock' and 'drawable' usage,
while fixing:
  - constructor: Add drawable null check -> IllegalArgumentException
  - setGLReadDrawable: Proper precondition checks
  - setGLDrawable: Proper precondition checks

Affected methods of mutable drawable for which we have to consider locking:
  - setGLReadDrawable
  - setGLDrawable
  - release
  - destroy
  - makeCurrent
 (commit: ad79bd0)

Bug 1054: Cleanup GLContext: 'lock' -> 'contextLock'
 (commit: dfb9ed4)

Bug 1054:  GLContext: makeCurrent() needs a null-check of [mutable] drawable; Review null checks and synchronization/locking.

'drawable' field of GLContextImpl is mutable via setGLDrawable(..),
which requires high-level locking as documented.

The required high-level locking allows us to _not_ add special
synchronization to this field (and drawableRead).

A simple null-check in makeCurrent() shall be sufficient,
plus ensuring mentioned high-level locking is applied.

GLContextImpl 'drawable' and 'drawableRead' synchronization:
  - commit ad79bd072b600a3f2416cc6f0c61e2925000069d check of null drawable is sufficient
  - Add GLAutoDrawable upstream-lock locking to:
    - AWT GLCanvas setupPrint/releasePrint
    - AWT GLJPanel (was missing)

Misc:
  - validate shared-context native-surface locking, throw exception if not successful
  - pixelDataEvaluated does not need to be synchronized, since it's being called while context is current, locking
  - GLDrawableHelper.recreateGLDrawable(..): Remove redundant glFinish() call
 (commit: 11347ad)

Bug 1055 - Access and query shared master GLContext in a deterministic fashion ; Don't use arbitrary shared context as 'master'.

GLContext* passes the shared-master to GLContextShareSet,
which only creates a sets of shared contexts
without differentiating the master context.

GLContext*'s shared-slave attempts to lock
the realized shared-master's surface at creation.

Currently only an arbitrary shared context is selected
due to the missing 'master' identity.
The arbitrary shared context's surface is locked
and its shared context handle used to create the slave context.

Lacking of using the user given shared-master
can lead to deadlock situations - and locking a 'wrong' surface.

+++

The patch:

  - Allows query the user given shared-master!

  - Use the user given shared-master for locking and it's
    context handle for the slave's creation.

  - The shared-context mapping maps each shared-master
    to a shared-slave within one shared-context-set,
    allowing deterministic and individual shared-master queries.
 (commit: 011e13e)

Bug 1054: Revert dfb9ed47ac6d8e85f6ae5fe166e7a6e28ca8ff83: Cannot change protected field name w/o breaking backward compatibility
 (commit: da16dfd)

Refine test cases for synchronous GLAutoDrawable display, swap-buffer and read-pixels (non-MSAA and MSAA) Bugs: 841, 975 and 1020
 (commit: 701e1e8)

Refine test cases for synchronous GLAutoDrawable display, swap-buffer and read-pixels (non-MSAA and MSAA) Bugs: 841, 975 and 1020

.. add Platform and GL info !
 (commit: 33072c5)

FBObject Cleanup: Add comments about maxSamples > 0 implies fullFBOSupport; bind(): Set dedicated read/write if fullFBOSupport

- Add comments about maxSamples > 0 implies fullFBOSupport

- bind(): Set dedicated read/write if fullFBOSupport
  as done in syncSamplingSink() and unbind()
 (commit: 7dbd7fc)

TestGLReadBuffer01GLJPanelAWT: Add keyFrame-wait after init
 (commit: 359ecbd)

Bug 1020 - First MSAA FBO frame on a mac osx nvidia card not antialiased

OSX/Nvidia's FBO needs to be cleared before blitting,
otherwise first MSAA frame lacks antialiasing.

GLFBODrawableImpl.initialize(..) can clear
GL.GL_COLOR_BUFFER_BIT and GL.GL_DEPTH_BUFFER_BIT, if used.

FBObject cannot clear the buffer(s) due to it's low-level API,
i.e. it cannot know when the first bind occurs _after_ user
completed FBO setup (attaching buffers).

Hence plain FBObject usage required manual injection
of glClear(..) after setup as demonstrated in GLJPanel.

We may need to elaborate in this case,
i.e. add an FBObject API entry like 'fbo.postInitNotify()'.
 (commit: ab07820)

TestSharedContextNewtAWTBug523: Fix junit usage ..
 (commit: 850cbdb)