Now with fixed mouse-over sensor, raycast support for bullet, and some extra python methods in that sensor. fixed dragging/damping conversion, so objects 'fall faster'.
http://www.continuousphysics.com/ftp/pu ... 05aug5.zip
enjoy
mouse-over sensor, raycast support for bullet
Hi Erwin. Seems to be coming along very nicely.
I had been getting instant crashes when trying
to start old games, right up to yesterdays build.
I hadent been able to discover what was causing it,
through subtraction. This build seems fine, although
old actions/armatures dont work. Probably time to
redo them anyway. The demo files are becoming
more realistic with just the occasional collision
between objects bouncing more than you would expect.
Great stuff.
Bogey.
I had been getting instant crashes when trying
to start old games, right up to yesterdays build.
I hadent been able to discover what was causing it,
through subtraction. This build seems fine, although
old actions/armatures dont work. Probably time to
redo them anyway. The demo files are becoming
more realistic with just the occasional collision
between objects bouncing more than you would expect.
Great stuff.
Bogey.
-
- Posts: 34
- Joined: Tue Jul 12, 2005 1:15 pm
yes, another big step!
However we need a test scenario everyone can compare with. For example I never played Jenaga with 3x1x1m Blocks
Once you scale down the scene (dont forget to Aplly Size/Rot...) everything breaks like in real life. Or at least similar. It is fun to reduce the Friction to almost nothing and see the stack collapsing like a fluid
Maybe time to modell a nice testscene like a Jenga stack on a Billardtable or such
Regards,
Carsten
However we need a test scenario everyone can compare with. For example I never played Jenaga with 3x1x1m Blocks
Once you scale down the scene (dont forget to Aplly Size/Rot...) everything breaks like in real life. Or at least similar. It is fun to reduce the Friction to almost nothing and see the stack collapsing like a fluid
Maybe time to modell a nice testscene like a Jenga stack on a Billardtable or such
Regards,
Carsten
-
- Site Admin
- Posts: 4221
- Joined: Sun Jun 26, 2005 6:43 pm
- Location: California, USA
small jenga stacks on a billardtable
There are a lot of ways to break a rigidbody dynamics system, and Bullet is no exception. It will require a lot of tweaking to get things right, both for the dynamics system writer and for the user.
Unfortunately small object sizes, lack of friction and damping, too much collision detail, large mass ratios, and many more issues can break the simulation.
First thing will be to get Bullet fully integrated in Blender and released for all platforms, Windows, Linux, FreeBSD etc. I don't expect small jenga stacks on a billardtable a realistic goal in the next half year
As a small example of commercial engine tweaking, did you read topic
http://www.continuousphysics.com/Bullet ... c.php?t=54 ,
Chris Butchers posting about Halo 2 and large mass ratios with Havok ?
Unfortunately small object sizes, lack of friction and damping, too much collision detail, large mass ratios, and many more issues can break the simulation.
First thing will be to get Bullet fully integrated in Blender and released for all platforms, Windows, Linux, FreeBSD etc. I don't expect small jenga stacks on a billardtable a realistic goal in the next half year
As a small example of commercial engine tweaking, did you read topic
http://www.continuousphysics.com/Bullet ... c.php?t=54 ,
Chris Butchers posting about Halo 2 and large mass ratios with Havok ?