Mantis Bug Tracker

View Issue Details Jump to Notes ] Issue History ] Print ]
IDProjectCategoryView StatusDate SubmittedLast Update
0006082opensim[REGION] OpenSim Corepublic2012-07-09 23:552012-07-19 23:25
ReporterKayaker Magic 
Assigned To 
PlatformLinuxOSopenSUSE 11.4 (x86_64)OS Version11.4
Product Version 
Target VersionFixed in Version 
Summary0006082: llWind points in a very different direction than the viewer wind
Description    I'm used to the SL llWind() returning a slightly different direction than the direction that the viewer wind points (where particles blow and flexies bend). On SL the two often look about 10 degrees apart, but generally point in the same direction.
    However on Open Simulator, the llWind(ZERO_VECTOR) seems to point in a completely random direction compared to the particle and flexy wind. Sometimes they almost point in the same direction, sometimes they point 180 degrees apart. I have not been able to see any pattern.
    I have verified that the problem happens on Open Simulator standalone and on 0.7.4 dev on osgrid. It happens with Imprudence, Phoenix or Firestorm viewers. I did all my testing in regular sized (256x256) regions.
Steps To Reproduce    Create a tall thin box,(.5,.5,4) mark it as flexy with softness 3, gravity 0 drag 1 wind 5. It will bend downwind. Place the following script in the box, it will generate a stream of particles that blow down wind. Note that the two point in the same direction, when the wind has given them time to settle down.
            list params=[ //make the smoke
                PSYS_SRC_BURST_PART_COUNT ,1,
                PSYS_PART_START_SCALE, <0.50,0.50,0>,
                PSYS_PART_START_COLOR, <0,0,0>

    Then create a box that is longest in the X direction (1.5,0.5,0.5). I like to put a notch in the negative X end by setting path cut end to 0.75. Then I can tell that the other end is the positive local X direction. Place the following script in this prim. It will start pointing in the direction of llWind.

        vector wnd=llVecNorm(llWind(ZERO_VECTOR));
        float ang = RAD_TO_DEG*llAtan2(wnd.y,wnd.x);
        if (ang<0.0) ang += 360.0; //display direction in +degrees

    Note that the viewer wind (particles and flexies) and the server wind (llWind) do not point in the same direction.
TagsNo tags attached.
Git Revision or version numberwhat does git stand for?
Run ModeStandalone (1 Region)
Physics EngineODE
Script Engine
EnvironmentMono / Linux64
Mono Version2.8
ViewerAny, Impudence, Phoenix..
Attached Filespng file icon Wind_003.png [^] (310,941 bytes) 2012-07-09 23:55
png file icon WindDir_001.png [^] (263,765 bytes) 2012-07-19 23:22

- Relationships

-  Notes
Talun (manager)
2012-07-11 01:03
edited on: 2012-07-11 02:28

Try creating a 16sqm plot and make sure all the prims are in that small parcel. Looks pretty close when you do that.

Update. wind values seems to be generated on a 16x16 grid for the region so is probably consistent in that area

Kayaker Magic (reporter)
2012-07-19 23:25

OK, I placed both prims on exactly the same location. Parcel size should not matter. Look at the WindDir_001.png image I uploaded. It shows the sever wind pointing at SSE while the viewer wind (flexies and particles) is pointing almost 180 degrees different at NNE. When the wind changes direction every few seconds, both the prims turn to a new uncorrelated direction. Perhaps you were lucky and they accidental pointed in the same direction? Wait a while and see what happens.

- Issue History
Date Modified Username Field Change
2012-07-09 23:55 Kayaker Magic New Issue
2012-07-09 23:55 Kayaker Magic File Added: Wind_003.png
2012-07-11 01:03 Talun Note Added: 0021783
2012-07-11 02:28 Talun Note Edited: 0021783 View Revisions
2012-07-19 23:22 Kayaker Magic File Added: WindDir_001.png
2012-07-19 23:25 Kayaker Magic Note Added: 0021844

Copyright © 2000 - 2012 MantisBT Group
Powered by Mantis Bugtracker