[Announcement] JMonkeyEngine 3.0 fully supports JOGL 2.0

[gouessej]

Hi

I spent 4 days to port JMonkeyEngine 3.0 to JOGL 2.0 and it is now operational:
http://ur1.ca/aonv9

A few features are missing including applets, NEWT, JInput For JogAmp and off-screen buffers.

I need to fix the build and the deployment, just copy the JARs containing the native libraries into /dist/opt/jogl and it will work.

The renderer based on the fixed pipeline and the other one based on the programmable pipeline work. They will replace the renderers based on our competitor (it's up to the core developers to decide, it's not up to me) or they will (probably) keep maintained as an extension of this engine by the JogAmp Foundation.

Long live JOGL 

[Riven]

Congrats on getting it to work.

Can you point out the advantages for the people using JME3?

[gouessej]

Thank you. You know that I don't want to start any flame war and it is difficult to answer your question without risking to do it because JMonkeyEngine 1, 2 and 3 primarily supported the competitor of JOGL and I'm the guy who drove the JOGL 1 renderer of JMonkeyEngine 2.0 more reliable (with the help of Mathias Henze).

My primary objective consists in doing something isofunctional, i.e something working as reliably and providing at least the same features than the existing renderers based on the competitor. "My" renderers will be maintained whatever the decision of the core developers and I don't want to harm people who want to go on using the existing renderers. Just look at the source code, I tried to avoid driving "my" renderers too much different of the existing ones.

In my humble opinion, JogAmp APIs deserve to be used in the most popular 3D engine written in Java at least as much as their competitors including AndroidGL. I am responsible for engine support at the JogAmp Foundation, I took 4 days off to work on this port.

If it was only up to me, Ardor3D and JMonkeyEngine 3 would be merged, JogAmp and its main competitor too. I hate effort duplication and such ports are quite boring. I wasted some time because I initialized ALUT (JOAL 1.1.3 OpenAL Utility) at the wrong moment, I forgot to set a flag to true to indicate whether the canvas is renderable and I used null terminated strings whereas it was not needed which broke 2 things in the shader handling. I would have preferred spending much time in looking for a new job and a new girlfriend or in working on TUER (which is 6 years old now).

When Renanse refused my help several years ago when I wanted to port JMonkeyEngine to JOGL, I asked myself another question:
Can they point out the advantages for the the people using JMonkeyEngine 2.0?
I was not thinking about JOGL.

If there is no need of having two very similar bindings for OpenGL, OpenAL, OpenCL, Open...I_don't_know_yet_L, I highly encourage the concerned people to work together. "The more the merrier"... not for me. I like computer science but I prefer eating some nice French food and making love (with beautiful French women if possible) rather than spending time on boring tasks... oh yeah I have to replace glGet... by gl.glGet... or gl.getGL2().glGet...

Ok I'm tired and I need to sleep a bit before my friend is on the air (radio).

Games published by our own members! Check 'em out!

[ra4king]

I actually (surprisingly) agree with gouessej on this issue: I hate effort duplication.

I don't see the need for many different libraries that overall achieve the same function....in the name of "competition" and "diversity". It only leads to fanbois, flame wars, and TONS of time wasted trying to see "which one is easier/better" for a certain project.

[StumpyStrust]

I think the problem is like why there is not one coding language that everything runs on. Some people think it should be done one way where others think it should be done another even if the results are the same.

I am also agree that since they are both just opengl bindings that there should be very little to disagree on but hey w/e I am not the one doing all the work.

I started with lwjgl and it has yet to lead me astray. As I always say "if it ain't broke, don't fix it." I have yet to find something I could not do with java/opengl so I am not going to change things. (switch to c++ or c#)

Edit:
Congratulations on doing what you did. I would not even know where to begin on doing something like this.

[gouessej]

I actually (surprisingly) agree with gouessej on this issue: I hate effort duplication.

Actually, you probably agree with me on some other things.

[size=5pt]I was sad to hear my friend on a Catholic radio known here to be against homosexual marriage... I'm atheist too[/size]

[nsigma]

I actually (surprisingly) agree with gouessej on this issue: I hate effort duplication.

I agree with him on the food and love  but not this necessarily.  Competition can be good, even in the open-source world; and projects trying different ways to achieve similar things is also good, if they learn from each other's successes and mistakes!

Right now, I'm close to considering a switch to JOGL to get access to NEWT - would love for LWJGL to gain some decent multi-window support! <hint>   

[princec]

Oh, the irony.

Cas

[gouessej]

Competition can be good, even in the open-source world; and projects trying different ways to achieve similar things is also good, if they learn from each other's successes and mistakes!

JogAmp and its main competitor share some source code but I'm the only one at the JogAmp Foundation that really feels this situation as a "competition". I don't think it is good in our case. JogAmp road map does not depend on its competitors, we have our own objectives.

[ra4king]

Oh, the irony.

Cas

What irony? O_o