away3d / awayphysics-core-fp11 Goto Github PK

If you create a compound shape and add a box shape to that compound shape.
If the box shape doesn't touch the center of the compound shape it stops a hinge from rotating.
Easier to show than expain.
http://www.shrewballooba.co.uk/ConstraintTest2.as

try either
modelCompound.addChildShape(modelPart1Shape, new Vector3D(0, 0, -900), new Matrix3D()); //rotates
//modelCompound.addChildShape(modelPart1Shape, new Vector3D(0, 0, -1000), new Matrix3D()); //sticks

Any chance of adding an Away version check to the physics.
maybe in AWPDynamicsWorld.getinstance
DEV branch

if (Away3D.MAJOR_VERSION ==4 && Away3D.MINOR_VERSION < 1)
throw new Error("Incorrect AWAY3D version "+Away3D.MAJOR_VERSION+"."+Away3D.MINOR_VERSION+ ". Use Away3D 4.1 or higher");

Need proper way to create code-controlled kinematic objects (elevators, moving platforms etc.)

CF_KINEMATIC_OBJECT not working

Using simple static objects (mass 0) with manually setting coordinates each frame, produces lots of errors - dynamic objects penetrate, jump around etc. Calculating and setting lineravelocity for moving object helps a bit, but still errors - especially when fps drops.

As I understand in original bullet, motion state is interpolated for kinematic objects.

This is a must have for many projects.

Needed both types of subg to work with bvh.

Hi All!

Currently I'm maintaining the CrossBridge SDK Community (updated) version (post FlasCC/Alchemy) which you can find here:
https://github.com/crossbridge-community/crossbridge

It would be really great if you could send pull requests regarding the bullet physics lib:
https://github.com/crossbridge-community/crossbridge-swc-bulletphysics

You can find new builds of the SDK here:
https://sourceforge.net/projects/crossbridge-community/

Also any other contribution is appreciated.

Best Regards!

Hi Li i have a new problem for you ... :)
Class AWPCollisionObject

Original code:

public function set skin(value:ObjectContainer3D):void
{
//
m_skin = value;
_originScale.setTo( m_skin.scaleX, m_skin.scaleY, m_skin.scaleZ );
}

Modified code:

public function set skin(value:ObjectContainer3D):void
{
//
m_skin = value;
//
if( m_skin != null )
_originScale.setTo( m_skin.scaleX, m_skin.scaleY, m_skin.scaleZ );
}

This is because in Games Maker the extra mesh class object contains all the physics object properties ( and more ). Whe these property are exported, this duplicate all the extra properties ( one, standard for the mesh, and one referenced by the body ) and in my case this cause a wrong result when the mesh are imported ... And because the object are the same before it is saved and after i need to set this properties to null ....

Ok Li ?

We're using AwayPhysics Dev Branch (to use some Away3D 4.1 features) and DebugDraw is not working.

See here
http://away3d.com/forum/viewthread/3146/

I recompiled the old Main branch to fix this, but Ive no idea how with alcehmy2 (DEV) branch.

Could you please compile with the (updateWheelTransform set to false) line
https://github.com/away3d/awayphysics-core-fp11/blob/dev/Bullet/AwayPhysics.c#L944

updateWheelTransform(j,false)

The problem happens much more if you use a higher step frequency (meaning the 3rd parameter to the away physics world step function is lower).

Some background:

The awayphysics collision callbacks are called from the awayphysics step function.

As suggested here: http://www.bulletphysics.org/mediawiki-1.5.8/index.php/Collision_Callbacks_and_Triggers
the awayphysics step function iterates over all of the contact manifolds.

The problem is that awayphysics doesn’t follow this additional advice from that link:

“This should be done during a simulation tick (substep) callback, because contacts might be added and removed during several substeps of a single stepSimulation call.”

So, what is happening is that if the collision occurs during one of those in between sub-steps, the collision callback won’t be called.

I initially reported this here: http://away3d.com/forum/viewthread/2997/ and JohnBrookes posted there that he was able to reproduce the problem.

by dev branch debugdraw 3d file format model not correct

I'm using: Adobe Air 3.8 Asc 2 or (Flex 3.7 without asc2) and ad-hoc build fails with error

Exception in thread "main" java.lang.OutOfMemoryError: Java heap space
at adobe.abc.GlobalOptimizer.createBlock(GlobalOptimizer.java:9952)
at adobe.abc.GlobalOptimizer.merge(GlobalOptimizer.java:9984)
at adobe.abc.GlobalOptimizer$InputAbc.readCode(GlobalOptimizer.java:1265
)
at adobe.abc.GlobalOptimizer$InputAbc.readBody(GlobalOptimizer.java:685)

at adobe.abc.GlobalOptimizer$InputAbc.readBodies(GlobalOptimizer.java:40
4)
at adobe.abc.LLVMEmitter.generateBitcode(LLVMEmitter.java:331)
at com.adobe.air.ipa.AOTCompiler.convertAbcToLlvmBitcodeImpl(AOTCompiler
.java:611)
at com.adobe.air.ipa.BitcodeGenerator.main(BitcodeGenerator.java:104)
Compilation failed while executing : ADT

APK setup creation FAILED.

The constraint ( hinge ) can be succesfully moved ( initially the constrained object moves a bit away from the constraint but... ) . So the constraint can now be placed at the right pinball palette position ! ( I have spent some times for work with the program interface, materials, material methods, Lua scripts, lights, camera, ecc but the phisycs remain a priority in Game Maker ! ) so, for now, thanks Li !
Some Games Maker images here:
https://www.facebook.com/pages/Estiva-Community/248102938603578?ref=hl

I think there is a bug in the collision dispatcher, I explain it in this thread http://away3d.com/forum/viewforum/45/

Good job and thanks!

"Update – Nov 22, 2011: Starting with Flash Player 11.2 and AIR 3.2, content targeting Flash Player 11 and AIR 3 (i.e., content using SWF version 13 and above) will not support the experimental Alchemy prototype."

http://blogs.adobe.com/flashplayer/2011/09/updates-from-the-lab.html

awayphysics crushs in AIR for IOS. We think it crushs when running alchemy. Is there any fix? or we just have to wait for alchemy to improve?

The current version of AwayPhysics doesn't support Away3D 4.1. I see there's a dev branch (probably work in progress?) for the 4.1 release but not sure if its stable. We want to use the new GPU particle system that comes with 4.1 for a demo we're working on but we also need physics for the simulation.

Do you have an estimated release date for 4.1 support?

There is a bug that make the wheels and the car body make strange movements when use a variable delta time.

Here is a demo: (takes a few seconds to load)
https://dl.dropbox.com/u/28132609/demo_car_away/index.html

Use arrows to control the car. You can see the bug when the car is turning.

Best regards

Hi Muzerly… I’m Luca and actually i’m working on Away Games Maker:

http://www.awaygamesmaker.altervista.org/

source:
http://79.44.103.235:8086/AwayGamesMaker/

At this time Games Maker are able to insert many kind of objects in to “one scene” . Make these object “physics” and simulate…
But, I have a problem with the Hinge constraint ( I don’t know with the others for now ). If I add one object, I make this object physics, I add a hinge constraint to this object
“ I add a motor to simulate for now because Games Maker use LUA ( script ) for control the scene objects”
And than I move this object “relocate” in to the scene …

         public function get pivotInA():Vector3D
         {
                // CModule.readFloat(pointer + 680);
                return _pivotInA;
         }

// The set properties function does not exist in to your source:

         public function set pivotInA( pivotInA:Vector3D ):void
         {
                _pivotInA = pivotInA;
                var vec1:AWPVector3 = new AWPVector3();
                vec1.sv3d = _pivotInA;

• pointer + 680 ( offset ??? ) unknow

• vec1 struct size = unknow

              CModule.writeBytes( pointer + 680, sizeof( vec1 )???, vec1 );
              CModule.free(vec1.pointer);
       }
  

Because Games Maker ( I don’t know if others prog. Or something else need to modify this value ) must be able to constantly manage scene objects…
Can you give me this data  or i need to “analize” the source 

PS: PhysX for linux can be downloaded from the NVidia site and if you navigate to the Games Maker folder, you can find the latest version…

Hi Muzerly and thanks anyway, Luca

I'm trying to modify awayphysics for my project, and I need to modify the crossbridge-compiled (or flascc-compiled) AwayPhysics.swc. In short, I want to add some API calls to check for collisions (i.e., expose btCollisionWorld::contactTest()), and maybe make "tick callbacks" more flexible.

So I'm trying to recompile AwayPhysics.swc. I see the command at the top of Bullet/AwayPhysics.c, but when I try it, I get the following error:

/var/folders/b1/pr8g96r90lj72kb9tv06nvhh0000gn/T//ccAv0jsV.o: error: undefined reference to '__ZTI23btPolyhedralConvexShape'

Perhaps I need to rebuild the libs Bullet/libbullet*.a myself? I don't know.

I'm using CrossBridge 1.0.1 on a Mac. If someone could help me out (and perhaps add some notes on building to the README.txt?) I'd much appreciate it.