| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
| |
Change terrain activation state to DISABLE_SIMULATION for better performance.
|
| |
|
| |
|
|
|
|
| |
assets weren't already in the cache. Comment cleanups.
|
|
|
|
| |
failed flag in BSPrim.ForceBodyShapeRebuild()
|
| |
|
| |
|
|
|
|
| |
linksets into own subclass. Will eventually add manual movement linkset subclass.
|
|
|
|
| |
OdePrim.BadMeshAssetCollideBits that monodevelop inserted automatically
|
|
|
|
| |
Fix spelling of collide, change to more self-documenting property BadMeshAssetCollideBits, add method doc, change to private to reduce code analysis complexity
|
|
|
|
|
|
| |
handling.
Refactor some names to make them available for the asset tracking and fetching.
|
|
|
|
|
|
| |
Various tweekings to avatar shape/mass/inertia/etc.
Remove change from avatar radius to diameter. But still the avatar sinks.
Collision_end now happens immediately rather than at the next subscription time.
|
|
|
|
|
|
| |
Change character scaling to represent size of capsule (diameter rather than radius)
Modify create capsule call to pass radius and height.
Eliminate errors when calculating shape inertia (should have some type checking).
|
|
|
|
|
| |
Add new function to ParameterDefn for calling BulletSimAPI to set values.
Tweaking to BSCharacter parameter setting to try and have avatars stand.
|
|
|
|
| |
initialized and setting of C++ parameters commented out. Comments and logging added.
|
|
|
|
|
|
|
|
| |
Add capsule shape to BSShapeCollection().
Remember last updated values so inter frame diffs can be computed.
Parameterize avatarStandingFriction and reduce to 10 from 999.
The latter high value made avatars very hard to push.
Set CCD parameters for prims and characters of specified.
|
|
|
|
| |
the defaul basic box so they don't go off world.
|
|
|
|
| |
assets behave like phantom by Nebadon request
|
|
|
|
| |
Signed-off-by: Melanie <melanie@t-data.com>
|
|
|
|
| |
Signed-off-by: Melanie <melanie@t-data.com>
|
| |
|
|
|
|
| |
as cubes.
|
|
|
|
| |
Bullet 2.81
|
| |
|
| |
|
|
|
|
| |
taint-time specification to new AddForce().
|
|
|
|
| |
more rigid. Remove mass calculation for terrain (it should stay a static object).
|
|
|
|
| |
improve interactions.
|
|
|
|
|
|
|
|
|
|
| |
Update BSDynamics to use same (don't want to delay updates til next taint-time.
Suppress queuing a taint update for position and orientation calls if value
does not change.
Move Bullet timing statistics call from C# back to C++ code.
Throttle taints per simulation step and add parameter to set.
By default, don't create hulls for physical objects. Add a
parameter to turn on and off.
|
|
|
|
|
|
|
|
| |
Properly remove and restore linkage constraints when upgrading
a prim's mesh to a hull.
Lots more debug logging.
Definitions and use of Bullet structure dumping.
Centralize detail logging so a Flush() can be added for debugging.
|
|
|
|
| |
No Avatar Avatar Collisions. This causes avatar to be able to walk through each other as if they're ghosts but still interact with the environment.
|
|
|
|
| |
'pushability' av_planted see OpenSimDefaults.ini. Use when you have unruly visitors that rudely push each other around. Still allows a small amount of movement based on the avatar movement PID controller settings. You can increase the spring tension in the PID controller and really prevent any movement if you would like.
|
|
|
|
|
| |
This introduces expected contact point hints to the ODE Collider to better determine when to throttle updates as excessive. This /should/ smooth physics objects out again, however, I cannot know every configuration of OpenSimulator, so I'm requesting that testers please examine this change on their build.
Thanks!
|
|
|
|
| |
to control the reporting of a new angular velocity. I think that this could be here for one of two reasons, 1. vehicles and llMoveToTarget with axis lock, or 2. To attempt to make things look more stable in the physics scene then they really are, however, this also really affects the angular velocity reporting negatively causing things to spin wildly and jump back into place repeatedly. To compromise, if the prim is a vehicle or is being used as a motor target, the original functionality is still applied. If that's not the case, angular velocity is reported with a linear velocity of 0.02m/step. To be clear on the effect of the physical world... When you push things, there's still a lag time where you walk into the object but once the object is in motion, it begins to move as you would expect so results in slightly more realistic motion.
|
| |
|
|
|
|
| |
and what is in OpenSimDefaults.ini. Comment and debug printout changes.
|
|
|
|
| |
children of BSPrim and BSCharacter.
|
|
|
|
| |
type changed.
|
|
|
|
| |
was using the child shape address at call time rather than the one created at taint time.
|
|
|
|
| |
no lag between what the vehicle code sees and what the physics engine is using.
|
| |
|
| |
|
| |
|
| |
|
|
|
|
| |
destroyed. This fixes many problems with physical linksets.
|
|
|
|
| |
keep the creation of constraints separate from runtime.
|
|
|
|
|
|
|
|
|
| |
use.
Added callbacks for shape and body changes in GetBodyAndShape() so the linkset
constraints can be picked up and restored. A better design might be to have
a "prim shape changed" event. Think about that.
Added constraint types to general constraint class.
|
| |
|
|
|
|
| |
going back now.
|
|
|
|
|
|
|
| |
pointers at runtime and at taint-time. Now passes the body into the
taint.
Vehicles zero inertia when active to eliminate Bullet's contribution
to vehicle motion.
|