I m currently thinking about Ogre integration. Must be a part of Gluon, or not ?

Look this ! http://www.ogre3d.org/forums/viewtopic.php?t=45709

So advantage is it's only a 3d engine, there is no dependecy, and it works well with Qt.
Now the question is: Should Gluon be a Big API for every games or a simple API to make 2d games.
It's not hard to use Gluon. We just have to create KGLOgreView and some other class like KGLOgreSettingsWidget.
Maybe , it's better to create another one lib called kogre... However KAL and KCL can be use perfectly with ogre!

Example :
I want to create a 2d games :
I need Gluon :
KGL for 2d graphics
KAL for sound
KCL for input.
I can use this lib by link my program with : -lGluon or -lkal -lkgl -lkcl

I want to create a 3d games:
I need Ogre!
kglOgre
kal
kcl

...
So the question is : Must Ogre be a part of Gluon, or not ?

Just a thought: Would it be possible to make "KGL3D" abstract enough so that other 3D engines can be used as backends in the future? Unless of course OGRE has become the sort of de facto standard in the open source community that it doesn't make sense to go that extra mile.

Either way, whether part of Gluon or as a side project, it would be nice to have 3D in Gluon.

OGRE would be awesome *but* I'd say it's more important to get something ready to use and functional. If you get that, you'll build up a community of contributors who can then help with the harder job of supporting 3d games.

If you aim for something big and complex right from the start, there's a much better chance we'll never see anything completed. We see this on the Blender Game Engine forums all the time. Everyone wants to make the biggest, best game ever. The only ones that ever get finished are the ones that set a fairly easy target and keep going until it's done. Then they set a bigger target.

Don't get me wrong. If Gluon had the ability to do 3d graphics, AI, logic etc, it would be awesome. I'm just concerned that with a small team it may be taking on too much.