From 70be669b0db3713117982dbeeadbb91ed670e901 Mon Sep 17 00:00:00 2001 From: Sven Gothel Date: Wed, 14 Feb 2001 12:54:54 +0000 Subject: rel-2-5-2-0-test3 --- README-Version-2-5-2.txt | 126 +++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 126 insertions(+) create mode 100644 README-Version-2-5-2.txt (limited to 'README-Version-2-5-2.txt') diff --git a/README-Version-2-5-2.txt b/README-Version-2-5-2.txt new file mode 100644 index 0000000..eaf5c30 --- /dev/null +++ b/README-Version-2-5-2.txt @@ -0,0 +1,126 @@ +GL4Java has it's new _legal_ name: + + OpenGL[tm] for Java[tm] + + It's abbreviation may be gl4java .. with the additional trademark hints ! + +Summary Version 2-5-2-0 testX: +=============================== + - jdk1.3 and jdk >=1.4 support while implementing the JAWT interface ! + - multi threading usage is more reliabel. + - code is now 64bit clean, 64bit prepared + - Using the new revision makes your existing code incompatible, + because: + - The GLUFunc.gluNew* etc. methods uses the long type for + the reference holder for quadric, tesselation, + .. (old: int type) ! + + This is, because this new revision is completly 64bit clean ! + All native reference holders are of type long ! + + - You did not encapsulate your OpenGL calls within + GLContext.gljMakeCurrent + + GLContext.gljFree + e.g. within the GLCanvas.display method: + glj.gljMakeCurrent(); // lock GL Context + gl.glBegin(); + gl.glEnd(); + glj.gljSwap(); + glj.gljFree(); // release GL Context + + All the demo's have been updated in the new test release, + to work with 2.5.2.0 ! + +The prerelease can be fetched: + + binary-packages: + http://www.jausoft.com/Files/Java/1.1.X/GL4Java/binpkg + + sources, doc's, demo's: + http://www.jausoft.com/Files/Java/1.1.X/GL4Java/archive + + The CVS repository is commited and tagged: rev-2-5-2-0-testX + http://gl4java.sourceforge.net/ + +Experience: +========== + Linux 2.4.1, XFree86 4.0.2, kde 2.1 beta2: + + Command-Line JDK (1.1-1.3, sun, blackdown, ibm(no jawt)): OK ! + Indirect DRI (Mesa-Soft 3.3): OK + NVidia 0.96: OK + See NVidia notes below ! + + Netscape + Netscape JVM: ERROR ! + does not work, because Netscape's JVM + implementation on linux looks like + not to handle long-types within the JNI API .. ?! + + Netscape + j2re-1_3_0_01: OK ! + Indirect DRI (Mesa-Soft 3.3): OK + NVidia 0.96: OK + See NVidia notes below ! + + Konqueror (2.1 beta2): + jdk 1.3-sun: ERROR + Applet runs, but nothing is visible ! + + jdk 1.3-ibm (no jawt): OK ! + + jdk 1.3-blackdown: OK ! + Indirect DRI (Mesa-Soft 3.3): OK + NVidia 0.96: OK + See NVidia notes below ! + + Win98/WinNT4.0(*): + + Command-Line JDK (1.1-1.3, sun): OK ! + Well, you have to start the java.exe from the + /jre/bin/java.exe + Otherwise, the jawt.dll is not found !! BUG in JDK !! + + IExplorer + MS-JVM buil 3802: OK ! + JDirect usage .. + + JRE 1.3* Plugin + ( IExplorer | Netscape ) OK ! + + Netscape + Netscape JVM OK ! + + (*) WinNT 4.0: + On my system (servicepack6, nvidia detonator3), + i cannot restart the applet ! + An read access to adress 0x000000 occures right from + NVidia's OGL implementation !! + Looks like another bug within NVidia ? + See NVidia notes below ! + Well, Win98 runs fine + + +NVidia Driver Note's: +==================== + +The Linux 0.96 driver, and may be the winnt4.0 drivers, +looks like not being able to handle THREAD's very well ! + +Example: + +JVM Process 1: + JVM Process 1 - Thread 1: + >MakeCurrent() + >Free() + + This Produces many .X.err logs regarding to new created drawables for + same thread !(?) + +If you use the same JVM process, and you start many applets -> native threads, +the GL context managment looks like very confused. +On Linux the GL rendering went slower and slower ! +But a restarted JVM is fast as its first time ! +Yes - GL4Java does destroy the glcontext +and finalizes well ! + +comments, ideas are very wellcome ! + +kind regards, sven + -- cgit v1.2.3