Will this be possible in future releases(2.42 perhaps)

[Icoxo]

1). set the limits of the constraints, like a hinge constraint it can't go past a certain angle
2). destroy them using a logic brick or python so you can have a connected set of bodies and then break the connections. Would be good for say, welded structures that break apart.
3). stack objects without seeing a tiny amount of space between the objects, like you're looking at a stack of boxes and there's a little space between the top of one and the bottom of the one on top of it.

A side note, somehow I have to log in twice to get logged in the site.

[Erwin Coumans]

1). set the limits of the constraints, like a hinge constraint it can't go past a certain angle

Probably in a future release.

2). destroy them using a logic brick or python so you can have a connected set of bodies and then break the connections. Would be good for say, welded structures that break apart.

removing constraints is already possible. There is some 'ragdoll' demo where you can shoot limbs and they break. See elysiun.

3). stack objects without seeing a tiny amount of space between the objects, like you're looking at a stack of boxes and there's a little space between the top of one and the bottom of the one on top of it.

if you choose 'box' for bounds, there should be no space. For convex, there is some space at the moment. This will be reduced in future releases.

A side note, somehow I have to log in twice to get logged in the site.

I had the same problem. I modified the admin settings for cookies. Can you try again, once ?

[Icoxo]

One more question, will the P2P and other constraints be accessable through the constraint logicbrick so those who don't know python can play with them.

Other then that thanks anyway, I think if you managed to get character support and hinge limits and the promised bugfixes and improved performance into the 2.42 game engine, then there would be absolutely no reason to use sumo anymore(which would be good since sumo physics only work well for balls)

[Etherton]

Can blender support native pluggins?
It is possible to integrate other physics engines with blender?
Are they license limitations?

[Erwin Coumans]

Can blender support native pluggins?
It is possible to integrate other physics engines with blender?
Are they license limitations?

I integrated almost every physics engine available behind the physics abstraction layer, including Havok 1.x,2.x,3.x, Novodex 1.x,2.x, MathEngine, Ipion, Dynamo, Ode, Bullet.

Blender is GPL so you won't be able to distribute it with a gpl incompatible license. In other words, you can't integrate Novodex, Havok, Newton, Tokamak, because they are closed source and non-gpl.

But you are allowed to do it all in-house, just don't distribute any binary/source.

[Erwin Coumans]

One more question, will the P2P and other constraints be accessable through the constraint logicbrick so those who don't know python can play with them.

Other then that thanks anyway, I think if you managed to get character support and hinge limits and the promised bugfixes and improved performance into the 2.42 game engine, then there would be absolutely no reason to use sumo anymore(which would be good since sumo physics only work well for balls)

Probably a logicbrick as well.
Thanks

[Icoxo]

This would be great news for new constraints to be accessable through logic bricks, I appreciate your efforts, the game engine is truly poised to be the best ever with 2.42.

[Etherton]

Blender is GPL so you won't be able to distribute it with a gpl incompatible license. In other words, you can't integrate Novodex, Havok, Newton, Tokamak, because they are closed source and non-gpl.

But you are allowed to do it all in-house, just don't distribute any binary/source.

I see, I was looking for an inexpensive modeler to write a level editor for my project and I thought blender would be a good choice.
My idea was to integrate Newton for physics, but given I would not be able to distribute a plugging it would be a waste of time.
It is there any other inexpensive modeler for an indie developer with a less restrictive license?
Thanks for your reply anyway you save me a great deal of time.

[Erwin Coumans]

I see, I was looking for an inexpensive modeler to write a level editor for my project and I thought blender would be a good choice. My idea was to integrate Newton for physics, but given I would not be able to distribute a plugging it would be a waste of time.
It is there any other inexpensive modeler for an indie developer with a less restrictive license? Thanks for your reply anyway you save me a great deal of time.

Ah level editor... Perhaps I'm adding some plugin infrastructure, which would allow doing this in the future. But for now, did you check out Milkshape? Also, if you just need a modeler, you can consider using Collada exporters to author some physics content.

There is upcoming support for Collada Physics, with several modelers exporting into this format. Maya, Max and Blender have basic Collada support, but only Maya has the latest 1.4 Collada Physics.

http://colladamaya.sourceforge.net for Maya exporter
http://sourceforge.net/projects/collada-dom for Collada importer

Max, Blender and XSI support for physics is upcoming.
So you can consider writing a Collada Physics importer/binding for Newton. I'm planning to do this for Bullet, and because of the physics abstraction layer, it would work for other physics engines too.

Also see collada.org (needs registration).

[Etherton]

I do not think I am ready for Collada just yet. I find too cumbersome.
I am thinking a proprietary format that could only be use to exports and import scene for me engine, kind of Gmax.

I was thinking of writing the editor myself but it is lot of works, so I thought I could take a short cut.
Since my friend and I are using blender to exports models we thought that a full level editor with model editing would be the next step.
But thanks anyway blender is a great tool.