aboutsummaryrefslogtreecommitdiffstatshomepage
Commit message (Collapse)AuthorAgeFilesLines
* Merge branch 'master' of ssh://opensimulator.org/var/git/opensimJustin Clark-Casey (justincc)2013-01-246-2/+4
|\
| * Merge branch 'master' of ssh://opensimulator.org/var/git/opensimRobert Adams2013-01-2317-144/+672
| |\
| * | BulletSim: pass up and report the real collision penetration.Robert Adams2013-01-236-2/+4
| | |
* | | Fix mono 2.4.3 build break by using CreateInstanceAndUnwrap 9 method call ↵Justin Clark-Casey (justincc)2013-01-241-0/+1
| |/ |/| | | | | deprecated in later .net versions
* | Merge branch 'master' into coopterminationJustin Clark-Casey (justincc)2013-01-2465-12747/+1925
|\ \ | |/
| * Merge branch 'master' of ssh://opensimulator.org/var/git/opensimJustin Clark-Casey (justincc)2013-01-2459-12676/+1675
| |\
| | * BulletSim: remove setting of vehicle InterpolationRotationalVelocity.Robert Adams2013-01-232-36/+34
| | | | | | | | | | | | | | | | | | This doesn't seem to help the vehicle stability. Rename vehicle internal variables adding a "V" or "W" so it is clear when coordinates are vehicle or world relative.
| | * BulletSim: small change to center-of-mass computation left out last commitRobert Adams2013-01-231-3/+8
| | |
| | * BulletSim: center-of-gravity linkset changes. Not working yet.Robert Adams2013-01-234-29/+9
| | | | | | | | | | | | | | | | | | Conflicts: OpenSim/Region/Physics/BulletSPlugin/BSPrim.cs
| | * BulletSim: working on COMRobert Adams2013-01-234-7/+36
| | |
| | * BulletSim: fix build break introduced by previous commitRobert Adams2013-01-231-5/+0
| | |
| | * BulletSim: remove the unused RestoreBodyDependencies used by linksetsRobert Adams2013-01-233-33/+2
| | | | | | | | | | | | and vehicles and clean up code by removing their kludgyness.
| | * Add the Avination physics raycast glue so Core Physics can implement raycastMelanie2013-01-233-13/+129
| | |
| | * Remove the return value from llGiveMoney and add llTransferLindenDollars. AlsoMelanie2013-01-233-33/+114
| | | | | | | | | | | | | | | make llGiveMoney async so the script thread is not held up waiting for comms to an external server.
| | * Merge branch 'master' of /home/opensim/var/repo/opensimBlueWall2013-01-2360-12599/+1566
| | |\
| | | * BulletSim: Tweeks to vehicle motion.Robert Adams2013-01-213-15/+11
| | | | | | | | | | | | | | | | | | | | Pass through old angular velocity making for smoother transitions. Remove some old kludges for angular motion (damping and rotvel suppression).
| | | * BulletSim: allow changing position and rotation of a child of a linksetRobert Adams2013-01-218-46/+112
| | | | | | | | | | | | | | | | | | | | without rebuilding the whole compound shape. Should make vehicles move smoother.
| | | * Have SOP and LSL_Api call the proper GetCenterOfMass and GetGeometricCenterRobert Adams2013-01-212-1/+11
| | | | | | | | | | | | | | | | | | | | routines on the physics engine. Won't make a difference for any existing scripts since ODE always returned Vector3.Zero.
| | | * BulletSim: remove unused MaxTaintsToProcessPerStep parameterRobert Adams2013-01-213-8/+5
| | | |
| | | * BulletSim: More aggressive as setting character velocity to zeroRobert Adams2013-01-205-91/+153
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | when should be standing. Modify angular force routines to be the same pattern as linear force routines. BulletSim vehicle turning is scaled like SL and is DIFFERENT THAN ODE!! Fix some bugs in BSMotor dealing with the motor going to zero. Add a bunch of parameters: MaxLinearVelocity, MaxAngularVelocity, MaxAddForceMagnitude, VehicleMaxLinearVelocity, VehicleMaxAngularVelocity, and most of the values are defaulted to values that are larger than in SL. Use the new parameters in BSPrim, BSCharacter and BSDynamic.
| | | * BulletSim: fix problem of avatar sliding very slowly occasionally after ↵Robert Adams2013-01-201-34/+41
| | | | | | | | | | | | | | | | stopping walking. Consolidate movement tests into the one prestep motion action
| | | * BulletSim: small fix making sure terrain height is calculatedRobert Adams2013-01-201-1/+3
| | | | | | | | | | | | | | | | properly if the vehicle moves during vehicle actions.
| | | * BulletSim: modify motors to return correction rather than current valueRobert Adams2013-01-205-38/+36
| | | | | | | | | | | | | | | | | | | | | | | | | | | | to better use them for incremental updates. Modify prim and character to use the new motors. Simplify the vehicle linear movement code to just update the velocity directly or the basic movement.
| | | * This updates prebuild to remove BulletSimN, implements the BulletSim API in ↵teravus2013-01-205-296/+637
| | | | | | | | | | | | | | | | BulletSPlugin using the BulletXNA Bullet physics engine. It also updates the BulletXNA library to be compatible with the changes. OpenSimDefaults has been updated to describe how to switch engines and terrain implementations.
| | | * * Dumping BulletSimNPlugin in favor of combining the APIteravus2013-01-2023-11752/+0
| | | |
| | | * * Tweak the BulletSimN API a bit.teravus2013-01-202-17/+18
| | | |
| | | * * Allow unsafe code with BulletSimteravus2013-01-201-0/+2
| | | |
| | | * IRCBridgeModule: optional agent-alertbox for IRC enabled Regions look in ↵PixelTomsen2013-01-195-273/+316
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | OpenSimDefaults.ini / section [IRC] http://opensimulator.org/mantis/view.php?id=6470 idea: https://github.com/ssm2017/IrcBridgeAlert
| | | * minor: adjust formatting in TESTING.txt to make headers more wiki-likeJustin Clark-Casey (justincc)2013-01-191-5/+5
| | | |
| | | * Check the existing ScenePresence.ParentPart to make sure we're not trying to ↵Justin Clark-Casey (justincc)2013-01-191-2/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | sit on a prim we're already sat upon, rather than looking up the part from scratch. An adaptation of commit 055b8a2 Having both ParentID and ParentPart references now is redundant. ParentID should probably be eliminated.
| | | * Assign the SmartThreadPool name in the constructorOren Hurvitz2013-01-194-5/+30
| | | | | | | | | | | | | | | | This is required because some threads are created in the constructor, so assigning the name afterwards would be too late.
| | | * Explicitly stop PollServiceRequestManager() rather than relying on its ↵Justin Clark-Casey (justincc)2013-01-192-5/+15
| | | | | | | | | | | | | | | | | | | | | | | | destructor. Hopes to address occasional shutdown failures from http://opensimulator.org/mantis/view.php?id=6503
| | | * New constants for llGetObjectDetailsTalun2013-01-192-0/+92
| | | | | | | | | | | | | | | | | | | | | | | | | | | | New constants for llGetObjectDetails OBJECT_CHARACTER_TIME, OBJECT_ROOT, OBJECT_ATTACHED_POINT, OBJECT_PATHFINDING_TYPE, OBJECT_PHYSICS, OBJECT_PHANTOM and OBJECT_TEMP_ON_REZ also Pathfining constants, 3 of which are used by llGetObjectDetails
| | * | Add additional return statusBlueWall2013-01-232-4/+23
| | | | | | | | | | | | | | | | Adding additional return status for JsonRpcMethod. Now returns true/false
| * | | New constants for llGetObjectDetailsTalun2013-01-192-0/+92
| | |/ | |/| | | | | | | | | | | | | | | | New constants for llGetObjectDetails OBJECT_CHARACTER_TIME, OBJECT_ROOT, OBJECT_ATTACHED_POINT, OBJECT_PATHFINDING_TYPE, OBJECT_PHYSICS, OBJECT_PHANTOM and OBJECT_TEMP_ON_REZ also Pathfining constants, 3 of which are used by llGetObjectDetails
| * | Mantis 6507 keys returned by llGetAgentList incorrect for llList2KeyTalun2013-01-191-2/+2
| | | | | | | | | | | | The type of the keys returned by llGetAgentList corrected to LSL_Key
| * | Fix use of scene debug commands when region is set to root or a specific ↵Justin Clark-Casey (justincc)2013-01-191-15/+12
| | | | | | | | | | | | region where there is more than one region on the simulator.
| * | Add "debug set set animations true|false" region console command.Justin Clark-Casey (justincc)2013-01-184-24/+48
| | | | | | | | | | | | | | | Setting this logs extra information about animation add/remove, such as uuid and animation name Unfortunately cannot be done per client yet
| * | Restore previous client AO behaviour by not allowing them to remove the ↵Justin Clark-Casey (justincc)2013-01-185-8/+27
| | | | | | | | | | | | | | | | | | | | | | | | | | | default animation but continue to allow scripts to do so. This keeps the fix from http://opensimulator.org/mantis/view.php?id=6327 and fixes the behaviour regression in http://opensimulator.org/mantis/view.php?id=6483 Animations may still exhibit different behaviour if both scripts and clients are adjusting animations. A change in the behaviour of client AO to not remove all animations may be a better long term approach.
| * | BulletSim: reduce jitter in avatar velocity when walking or flying.Robert Adams2013-01-182-2/+9
| | | | | | | | | | | | | | | | | | | | | OpenSimulator is VERY sensitive to changes in avatar velocity and will send an avatar update message when velocity changes more than 0.001m/s. This significantly reduces the number of avatar update messages by smoothing the avatar velocity returned by Bullet.
| * | BulletSim: add logic to turn off pre-step actions when object goesRobert Adams2013-01-181-0/+21
| |/ | | | | | | | | non-active. This turns off 'setForce', 'setTorque' and 'moveToTarget' when the object is selected or made non-physical.
| * Merge branch 'master' of ssh://opensimulator.org/var/git/opensimRobert Adams2013-01-174-21/+6
| |\
| * | BulletSim: Add one function that all actors who act on the physicalRobert Adams2013-01-175-24/+31
| | | | | | | | | | | | | | | can use to know if the object is currently active. Code cleaning including use of Util.ClampV function.
| * | Add utility function to clamp a vector to a maximum magnitude.Robert Adams2013-01-171-0/+12
| | |
* | | Add information on ScriptStopStrategy to [XEngine] in OpenSimDefaults.ini ↵Justin Clark-Casey (justincc)2013-01-232-13/+31
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | and OpenSim.ini.example. Default remains abort. This setting controls whether scripts are stopped by aborting their threads externally (abort) or by co-operative checks from the compiled script (co-op) co-op should be more stable but this option is experimental. If moving from co-op to abort, existing script DLLs will need to be recompiled. This currently can only be done manually, either by setting DeleteScriptsOnStartup = true for one run or by deleting the script DLL* files in bin/ScriptEngines/<region-id>/ One can move from co-op back to abort without recompilation, but reverting back to co-op again will need script recompile
* | | Pass narrower WaitHandle rather than EventWaitHandle as co-op termination ↵Justin Clark-Casey (justincc)2013-01-235-6/+6
| | | | | | | | | | | | | | | | | | wait handle to script APIs. APIs don't need to reference any methods on EventWaitHandle
* | | Improve logging by making it clearer which script is failing if an assembly ↵Justin Clark-Casey (justincc)2013-01-233-28/+31
| | | | | | | | | | | | | | | | | | | | | | | | fails to load. Moves the noise co-op start/stop debug log messages to only display if xengine debug level >= 1 Logs which stop strategy is being used (abort or co-op) Adjusts some other logging to remove not very useful stuff
* | | If ScriptStopStrategy hasn't been set to co-op in [XEngine] config, then ↵Justin Clark-Casey (justincc)2013-01-236-27/+88
| | | | | | | | | | | | | | | | | | | | | | | | | | | continue to generate C# that is functionality identical to historical generation This is to eliminate disruption until co-op termination has been well-tested. In non co-op mode, XEngine will continue to load DLLs of the existing Script class and the new XEngineScript class. Moving to co-op mode still requires existing script DLL deletion to force recompilation, either manually or by setting DeleteScriptsOnStartup = true for one run. This change also means that scripts which fail to initialize do not still show up as running scripts.
* | | Add co-op termination regression test for infinite recursive manual call on ↵Justin Clark-Casey (justincc)2013-01-221-0/+22
| | | | | | | | | | | | | | | | | | event function. Such code would normally terminate quickly with a stack overflow exception anyway.
* | | Set script delay factor to 0 in co-op termination testsJustin Clark-Casey (justincc)2013-01-221-0/+3
| | | | | | | | | | | | | | | This is to ensure loops aren't actually terminating from a wait on an LSL function. This was not the case with any of the existing tests.