aboutsummaryrefslogtreecommitdiffstatshomepage
path: root/OpenSim/Region/Framework/Scenes/KeyframeMotion.cs (follow)
Commit message (Collapse)AuthorAgeFilesLines
* shut up some pesty warningsUbitUmarov2018-01-151-12/+12
|
* KFM: changes on update method implies other changes. Still not that good. ↵UbitUmarov2017-07-171-13/+25
| | | | Note: objects motion timing can't be done using timers. On windows at least, those can have jitter of tens of ms
* store kfm on inventory. this may still be a bit confusing on rezUbitUmarov2017-07-161-5/+6
|
* don't start moving KFM things on region start that where stopped on last ↵UbitUmarov2017-07-161-2/+2
| | | | shutdown
* trigger moving_end also KFM stop and pauseUbitUmarov2017-07-161-0/+2
|
* make KFM use scheduled terse updates. This may increase some rubber band, ↵UbitUmarov2017-07-161-12/+14
| | | | but is needed, and should work better with recent changes on sop terseupdates control
* mantis 8130: improve keyframes motionUbitUmarov2017-04-041-70/+63
|
* Massive tab and trailing space cleanupMelanie Thielker2017-01-051-4/+4
|
* Merge of ubitworkvarnew with opensim/master as of 20150905.Robert Adams2015-09-081-1/+1
|\ | | | | | | | | | | | | | | This integrates the OpenSim refactoring to make physics, etc into modules. AVN physics hasn't been moved to new location. Does not compile yet. Merge branch 'osmaster' into mbworknew1
| * More namespace and dll name changes. Still no functional changes.Diva Canto2015-08-301-1/+1
| |
| * Renamed the namespaces tooDiva Canto2015-08-301-1/+1
| |
* | seems to compile ( tests comented out)UbitUmarov2015-09-021-118/+10
| |
* | bad merge?UbitUmarov2015-09-011-33/+189
|\ \ | |/ |/|
| * bug fixUbitUmarov2014-10-261-1/+6
| |
| * fix positions on same frames transitionsUbitUmarov2014-09-191-8/+3
| |
| * remove creative hack and try to fixUbitUmarov2014-08-131-100/+92
| |
| * Merge branch 'avination-current'Melanie Thielker2014-07-191-15/+64
| |\ | | | | | | | | | | | | Conflicts: OpenSim/Region/CoreModules/Framework/InventoryAccess/InventoryAccessModule.cs
| | * Merge branch 'avination-current' of ssh://3dhosting.de/var/git/careminster ↵UbitUmarov2013-09-191-1/+1
| | |\ | | | | | | | | | | | | into avination-current
| | | * Fix the root cause of keyframe motion region crossing offset - UpdateSceneObjectMelanie2013-08-251-21/+1
| | | | | | | | | | | | | | | | was being called twice on each crossing.
| | | * * This fixes the border crossing offsets by storing the final keyframe ↵teravus2013-08-241-2/+22
| | | | | | | | | | | | | | | | location in the hijacked variable KeyFrame.AngularVelocity. When steps in OnTimer <= 0.0, normalize the final position by Constants.RegionSize and move the object there. The hack here is KeyFrame.AngularVelocity probably isn't the right name for this variable because it's the un-mucked with keyframe position. When you determine the feasibility of changing the name without affecting the serialization of existing objects in world... It's simply a name change to KeyFrame.FinalPosition or something proper.
| | | * Revert "* This increases accuracy when border crossing (the reason is.. ↵teravus2013-08-241-7/+7
| | | | | | | | | | | | | | | | | | | | | | | | the border crossing code will use velocity to predict where the object should be, so setting it to zero. It still looses about 0.0045 per loop." This reverts commit 55400ff7be55b1c8dbededca68e6fce42cd6ce0f.
| | | * * This increases accuracy when border crossing (the reason is.. the border ↵teravus2013-08-231-7/+7
| | | | | | | | | | | | | | | | crossing code will use velocity to predict where the object should be, so setting it to zero. It still looses about 0.0045 per loop.
| | * | keyframe: add more locks, change timming stop, added Suspend and resumeUbitUmarov2013-06-061-15/+64
| | |/
| | * Stop sending velocity to avoid snap-backMelanie2013-05-311-1/+1
| | |
| | * Make Keyframe motion cross regions againMelanie2013-03-031-0/+1
| | |
| | * Revert "Make keyframes use the sim's frame timer"Melanie2013-02-121-8/+15
| | | | | | | | | | | | This reverts commit e85a6237bfc0f00de4a183e29e515fa5baf1aa7f.
| | * Revert "Push updates from keyframe directly to the front of the output queue ↵Melanie2013-02-121-6/+6
| | | | | | | | | | | | | | | | | | rather" This reverts commit 04235e58e87ae42617111cad2884e42785914d4e.
| | * Push updates from keyframe directly to the front of the output queue ratherMelanie2013-02-121-6/+6
| | | | | | | | | | | | than through the update system.
| | * Make keyframes use the sim's frame timerMelanie2013-02-111-15/+8
| | |
| | * Refactor KeyframeMotion to use one timer class per sceneMelanie2013-02-111-20/+75
| | |
| | * Change keyframe motion to use a single timer for all objects. This is requiredMelanie2013-01-301-236/+229
| | | | | | | | | | | | | | | to prevent slippage between objects that are meant to move synchronously or keep their relative positions/rotations.
| | * Send moving_end event to scripts when keyframed motion ends.Melanie2013-01-271-0/+10
| | |
| | * go back bypassing sog updates control (test)UbitUmarov2013-01-111-12/+12
| | |
| | * same for AngularVelocity. Use normal terse updates in place of sendingUbitUmarov2013-01-111-10/+18
| | | | | | | | | | | | imediatly. If that's good for physics, needs to be good for this
| | * keyframe. Don't use group UpdateRotation since this enqueues a terseUbitUmarov2013-01-111-2/+6
| | | | | | | | | | | | update and we are sending them imediatly
| | * normalize quaternion.Slerp outputsUbitUmarov2012-10-101-0/+1
| | |
| | * move keyframemotion.copy from sop.copy to sog.copy, where there isUbitUmarov2012-08-301-2/+2
| | | | | | | | | | | | newgroup information avaiable.
| | * [possible still bad] make use of keyframemotion.copy on sop.copy, replacingUbitUmarov2012-08-301-25/+33
| | | | | | | | | | | | | | | fromdata(seralize). for now its called with null group since sop.copy() hasn't usable new group information, so for copied keyframes be fully operational UpdateSceneObject(newgroup) needs to be called on them.
| | * A few more changes to keyframesUbitUmarov2012-08-291-61/+113
| | |
| | * [Potentially broken] keyframes changes.. since it's there, use timer forUbitUmarov2012-08-291-115/+165
| | | | | | | | | | | | crossing retries and not still another thread, etc...
| | * [possible still very broken] mess around keyframes. timer eventsUbitUmarov2012-08-281-103/+265
| | | | | | | | | | | | threads overlaps, some null objects exceptions, region crossing...
| | * fix incoerence btw KFM_TRANSLATION and ROTATION LSL constants and internalUbitUmarov2012-08-261-2/+2
| | | | | | | | | | | | DataFormat enum, using values from the KFM constants
| | * Make llSetKeyframedMotion prims waut at the borders of down sims and try againMelanie2012-02-261-0/+12
| | | | | | | | | | | | periodically.
| | * Implement proper selection behaviorMelanie2012-02-261-0/+31
| | |
| | * Fix deserialization of Buoyancy, Force and Torque. Remove debug from the newMelanie2012-02-261-1/+0
| | | | | | | | | | | | code.
| | * Serialize Keyframe motion for region crossingsMelanie2012-02-261-3/+68
| | |
| | * Implement llSetKeyframedMotion. No persistence, no region crossing. Yet.Melanie2012-02-261-0/+315
| |
| * * This fixes the border crossing offsets by storing the final keyframe ↵teravus2013-09-161-2/+22
| | | | | | | | | | | | location in the hijacked variable KeyFrame.AngularVelocity. When steps in OnTimer <= 0.0, normalize the final position by Constants.RegionSize and move the object there. The hack here is KeyFrame.AngularVelocity probably isn't the right name for this variable because it's the un-mucked with keyframe position. When you determine the feasibility of changing the name without affecting the serialization of existing objects in world... It's simply a name change to KeyFrame.FinalPosition or something proper. (cherry picked from commit e0399ccaec68889c12e4679b4d62142b49b379df)
* | Fix multiple moving_end and add moving_start to llSetKeyframedMotionJeff Kelley2015-08-201-10/+15
| | | | | | | | Signed-off-by: Melanie Thielker <melanie@t-data.com>
* | Close streams immediately when we finish using themOren Hurvitz2014-07-211-21/+25
| |