Edge Rotate Function

heyas;

ged wrote a nifty little function that rotates an object on its z edge. which will no doubt work for my flappy flexi wings.... ::has to get back to that sometime soon!:: but, i thought maybe i could expand it and... well, make it do more.

so i did, and this is what i came up with. HOWEVER, i only works if the prim is not rotated, or sometimes if it is only rotated on one axis. once it starts getting rotated more, simple x rotations start getting translated into y rotations and other crap which i just can't handle.
if anybody wants to tinker with it and make it work perfectly all the time -- good luck! but that would be nice.

//--edge rotate originally by ged
//--expanded by bloodsong
//--not very reliable unless prims aren't rotated much to start

//--side constants.  pos is towards arrow point, neg is towards arrow butt
integer  X_POS = 1;  
integer  X_NEG = 2;
integer  Y_POS = 3;
integer  Y_NEG = 4;
integer  Z_POS = 5;
integer  Z_NEG = 6;

//--axis constants
integer X_AXIS = 1;
integer Y_AXIS = 2;
integer Z_AXIS = 3;

//--RotFromEdge: (angle, edge, axis)
//-- input your degrees in normal everyday degrees
//-- input what edge you want the axis on
//-- and which axis you want to rotate about
RotFromEdge(float fAngle, integer iSide, integer iAxis)
{
    //--part one initial setup
    vector size = llGetScale(); //--gets length of sides to find edges.
    vector currentRot = llRot2Euler(llGetLocalRot());
    
    rotation desiredRot;
    vector rotOriginBefore;
    
    //--adjust for different axes and edges
    if(iAxis == X_AXIS)
    {
        currentRot.x = fAngle*DEG_TO_RAD;
        desiredRot = llEuler2Rot(currentRot);
    }
    else if (iAxis == Y_AXIS)
    {
        currentRot.y = fAngle*DEG_TO_RAD;
        desiredRot = llEuler2Rot(currentRot);
    }        
    else if (iAxis == Z_AXIS)
    {
        currentRot.z = fAngle*DEG_TO_RAD;
        desiredRot = llEuler2Rot(currentRot);
    }    
    
    
    if(iSide == X_POS)
    {  
      rotOriginBefore = (llRot2Fwd(llGetLocalRot()) * size.x / 2.0);
    }
    else if (iSide == Y_POS)
    {
     rotOriginBefore = (llRot2Left(llGetLocalRot()) * size.y / 2.0);
    }
    else if (iSide == Z_POS)
    {
        rotOriginBefore = (llRot2Up(llGetLocalRot()) * size.z / 2.0);
    }  
    else if (iSide == X_NEG)
    {
      rotOriginBefore = (llRot2Fwd(llGetLocalRot()) * (size.x / -2));    
    }  
    else if (iSide == Y_NEG)
    {
      rotOriginBefore = (llRot2Left(llGetLocalRot()) * (size.y / -2));    
    }
    else if (iSide == Z_NEG)
    {
      rotOriginBefore = (llRot2Up(llGetLocalRot()) * (size.z / -2));    
    }
    
    //--part two, more calculations
    rotation newLocalRot = desiredRot * llGetLocalRot();
    vector rotOriginAfter;
    
    //--change from defaults, if needed
    if(iSide == X_POS)
    {
        rotOriginAfter = (llRot2Fwd(newLocalRot) * size.x / 2.0);
    }
    else if (iSide == Y_POS)
    {
        rotOriginAfter = (llRot2Left(newLocalRot) * size.y / 2.0);
    }
    else if (iSide == Z_POS)
    {
        rotOriginAfter = (llRot2Up(newLocalRot) * size.z / 2.0);
    }
    else if (iSide == X_NEG)
    {
        rotOriginAfter = (llRot2Fwd(newLocalRot) * (size.x / -2));
    }
    else if (iSide == Y_NEG)
    {
      rotOriginAfter = (llRot2Left(newLocalRot) * (size.y / -2));    
    }
    else if (iSide == Z_NEG)
    {
      rotOriginAfter = (llRot2Up(newLocalRot) * (size.z / -2));    
    }


    //--part 3, apply the rotation
   llSetPrimitiveParams([PRIM_POSITION, llGetLocalPos() + rotOriginBefore - rotOriginAfter, PRIM_ROTATION, newLocalRot]);

}



default
{
    touch_start(integer num) 
    {
//--this is just a test to see if it works.  this loops it around in a complete circle.
      integer n;
      
      for (n=0; n<12; ++n)
      {
          RotFromEdge(30.0, Z_NEG, X_AXIS);
      }
      
    }
}

The main reason I never expanded my code snippet was that it's very difficult to make generalizable. Sizes in the X, Y, and, Z dimensions are no longer accurate, if you have prim cuts / twists / skews, etc.

Calculating prim dimensions and the relative change of the "prim center" based on the primitive parameters -- now, that would be the true challenge.

heh;

well, if i could cut the prim in half, i wouldn't have to rotate it from the edge now, would i? ;) but yeah, finding an edge on a mutated prim would be a pain. and calculating it from cuts and twists and... WAY too much math for me!

heyas;

okay, i found a MAJOR flaw in my function, which has to do with the start values for calculating the rots. ged just used 0's on his non-active axes, and i thought that would only work if the object wasn't rotated to start. and that i would have to start with the current rotation. well, turns out i was wrong.

only one line change:
vector currentRot = <0.0,0.0,0.0>;


now it works, no matter how you rotate the fool prim to start. yay!

the bad news is, it TOTALLY fails to work right if the prim is a child of anything. ::sob::

also as noted above, not 100% accurate for figuring 'edge' of cut/twisted/sheared/etc prims.

//--edge rotate originally by ged 
//--expanded by bloodsong 
//--not very reliable unless prims aren't rotated much to start 

//--side constants.  pos is towards arrow point, neg is towards arrow butt 
integer  X_POS = 1;   
integer  X_NEG = 2; 
integer  Y_POS = 3; 
integer  Y_NEG = 4; 
integer  Z_POS = 5; 
integer  Z_NEG = 6; 

//--axis constants 
integer X_AXIS = 1; 
integer Y_AXIS = 2; 
integer Z_AXIS = 3; 

//--RotFromEdge: (angle, edge, axis) 
//-- input your degrees in normal everyday degrees 
//-- input what edge you want the axis on 
//-- and which axis you want to rotate about 
RotFromEdge(float fAngle, integer iSide, integer iAxis) 
{ 
    //--part one initial setup 
    vector size = llGetScale(); //--gets length of sides to find edges. 
    vector currentRot = <0.0,0.0,0.0>; 
     
    rotation desiredRot; 
    vector rotOriginBefore; 
     
    //--adjust for different axes and edges 
    if(iAxis == X_AXIS) 
    { 
        currentRot.x = fAngle*DEG_TO_RAD; 
        desiredRot = llEuler2Rot(currentRot); 
    } 
    else if (iAxis == Y_AXIS) 
    { 
        currentRot.y = fAngle*DEG_TO_RAD; 
        desiredRot = llEuler2Rot(currentRot); 
    }         
    else if (iAxis == Z_AXIS) 
    { 
        currentRot.z = fAngle*DEG_TO_RAD; 
        desiredRot = llEuler2Rot(currentRot); 
    }     
     
     
    if(iSide == X_POS) 
    {   
      rotOriginBefore = (llRot2Fwd(llGetLocalRot()) * size.x / 2.0); 
    } 
    else if (iSide == Y_POS) 
    { 
     rotOriginBefore = (llRot2Left(llGetLocalRot()) * size.y / 2.0); 
    } 
    else if (iSide == Z_POS) 
    { 
        rotOriginBefore = (llRot2Up(llGetLocalRot()) * size.z / 2.0); 
    }   
    else if (iSide == X_NEG) 
    { 
      rotOriginBefore = (llRot2Fwd(llGetLocalRot()) * (size.x / -2));     
    }   
    else if (iSide == Y_NEG) 
    { 
      rotOriginBefore = (llRot2Left(llGetLocalRot()) * (size.y / -2));     
    } 
    else if (iSide == Z_NEG) 
    { 
      rotOriginBefore = (llRot2Up(llGetLocalRot()) * (size.z / -2));     
    } 
     
    //--part two, more calculations 
    rotation newLocalRot = desiredRot * llGetLocalRot(); 
    vector rotOriginAfter; 
     
    //--change from defaults, if needed 
    if(iSide == X_POS) 
    { 
        rotOriginAfter = (llRot2Fwd(newLocalRot) * size.x / 2.0); 
    } 
    else if (iSide == Y_POS) 
    { 
        rotOriginAfter = (llRot2Left(newLocalRot) * size.y / 2.0); 
    } 
    else if (iSide == Z_POS) 
    { 
        rotOriginAfter = (llRot2Up(newLocalRot) * size.z / 2.0); 
    } 
    else if (iSide == X_NEG) 
    { 
        rotOriginAfter = (llRot2Fwd(newLocalRot) * (size.x / -2)); 
    } 
    else if (iSide == Y_NEG) 
    { 
      rotOriginAfter = (llRot2Left(newLocalRot) * (size.y / -2));     
    } 
    else if (iSide == Z_NEG) 
    { 
      rotOriginAfter = (llRot2Up(newLocalRot) * (size.z / -2));     
    } 


    //--part 3, apply the rotation 
   llSetPrimitiveParams([PRIM_POSITION, llGetLocalPos() + rotOriginBefore - rotOriginAfter, PRIM_ROTATION, newLocalRot]); 

} 



default 
{ 
    touch_start(integer num) 
    { 
//--this is just a test to see if it works.  this loops it around in a complete circle. 
      integer n; 
       
      for (n=0; n<12; ++n) 
      { 
          RotFromEdge(30.0, Z_NEG, X_AXIS); 
      } 
       
    } 
} 

okay, after SCOURING the forums here for anything to do with rotations, i found this Very Important Bit of Information:
/54/c1/94705/1.html

in short, llsetlocalrot and llsetprimparameters is bugged. now why didn't somebody just SAY so? come to think of it, why doesnt the WIKI say so? this seems a crucial workaround, here. nobody knows it???



okay, two functions here, which is probably overkill if all you want to do is rotate a flexiprim on its stable edge.

RotFromEdge, which works for single, root, or child prims but NOT on attachments (actually, a root on an attachment should work)

AttachedRotFromEdge, which works for child prims in an attachment, but you must pass it the root's local rotation.

enjoy!

//--ged larsson's edge rotation script, expanded to a function
//--to rotate on any edge/axis combination
//--fixed for child/attached prims with info from lex neva
//--compiled by bloodsong termagant

//--side constants.  pos is towards arrow point, neg is towards arrow butt
integer  X_POS = 1;  
integer  X_NEG = 2;
integer  Y_POS = 3;
integer  Y_NEG = 4;
integer  Z_POS = 5;
integer  Z_NEG = 6;

//--axis constants
integer X_AXIS = 1;
integer Y_AXIS = 2;
integer Z_AXIS = 3;

//--RotFromEdge: (angle, edge, axis)
//-- input your degrees in normal everyday degrees
//-- input what edge you want the axis on
//-- and which axis you want to rotate about
RotFromEdge(float fAngle, integer iSide, integer iAxis)
{
    //--part one initial setup
    vector size = llGetScale(); //--gets length of sides to find edges.
    //--NOTE: does not work for cut or twisted prims
    vector currentRot = <0.0,0.0,0.0>; //do not use the local rot! start at zero to add to current rotation (multiply, do weird quarternion math with, whatever)
    
    rotation desiredRot;  //--we will build this from our single angle
    vector rotOriginBefore;  //--starting origin position rotation
    
    //--adjust for different axes and edges
    //--insert our angle into the desired rotation
    if(iAxis == X_AXIS)
    {
        currentRot.x = fAngle*DEG_TO_RAD;
        desiredRot = llEuler2Rot(currentRot);
    }
    else if (iAxis == Y_AXIS)
    {
        currentRot.y = fAngle*DEG_TO_RAD;
        desiredRot = llEuler2Rot(currentRot);
    }        
    else if (iAxis == Z_AXIS)
    {
        currentRot.z = fAngle*DEG_TO_RAD;
        desiredRot = llEuler2Rot(currentRot);
    }    
    
    
    //--this gets the rotation times half (or negative half) side of the thing.
    //--this is ged's part, i have no idea what it really means :X
    if(iSide == X_POS)
    {  
      rotOriginBefore = (llRot2Fwd(llGetLocalRot()) * size.x / 2.0);
    }
    else if (iSide == Y_POS)
    {
     rotOriginBefore = (llRot2Left(llGetLocalRot()) * size.y / 2.0);
    }
    else if (iSide == Z_POS)
    {
        rotOriginBefore = (llRot2Up(llGetLocalRot()) * size.z / 2.0);
    }  
    else if (iSide == X_NEG)
    {
      rotOriginBefore = (llRot2Fwd(llGetLocalRot()) * (size.x / -2));    
    }  
    else if (iSide == Y_NEG)
    {
      rotOriginBefore = (llRot2Left(llGetLocalRot()) * (size.y / -2));    
    }
    else if (iSide == Z_NEG)
    {
      rotOriginBefore = ( (size.z / -2) * llRot2Up(llGetLocalRot()));    
    }
    
    //--part two, more calculations
    //--our new rotation will be our desired change mixed with the current rotation
    rotation newLocalRot =  desiredRot * llGetLocalRot(); // llGetRot(); //--not llGetLocalRot() *
    vector rotOriginAfter; //--as it rotates, the origin will change to something new
    
    //--change for different edges
    //--same as before, but with our new rotation
    if(iSide == X_POS)
    {
        rotOriginAfter = (llRot2Fwd(newLocalRot) * size.x / 2.0);
    }
    else if (iSide == Y_POS)
    {
        rotOriginAfter = (llRot2Left(newLocalRot) * size.y / 2.0);
    }
    else if (iSide == Z_POS)
    {
        rotOriginAfter = (llRot2Up(newLocalRot) * size.z / 2.0);
    }
    else if (iSide == X_NEG)
    {
        rotOriginAfter = (llRot2Fwd(newLocalRot) * (size.x / -2));
    }
    else if (iSide == Y_NEG)
    {
      rotOriginAfter = (llRot2Left(newLocalRot) * (size.y / -2));    
    }
    else if (iSide == Z_NEG)
    {
      rotOriginAfter = ( (size.z / -2) * llRot2Up(newLocalRot));    
    }


    //--part 3, apply the rotation
    //--but FIRST... find out if we are attached (in which case, this wont work), or a child, in which case, we need to divide by the root rot

   if(llGetAttached() != 0)
   {//--need a new function for this, with the roots rotation passed to it
       llOwnerSay("Error: Edge Rotate does not work on attachments.");
       return;
   }
   
   if(llGetLinkNumber() > 1)
   {//--if not lone (0) or root (1), use this formula
      rotation R = llGetRootRotation();
      llSetPrimitiveParams([PRIM_POSITION, llGetLocalPos() + rotOriginBefore - rotOriginAfter, PRIM_ROTATION, newLocalRot/R]);       
      return;
   }
   //--else it'll work fine.  use normal.
   llSetPrimitiveParams([PRIM_POSITION, llGetLocalPos() + rotOriginBefore - rotOriginAfter, PRIM_ROTATION, newLocalRot]);

}



default
{
 touch_start(integer num) 
 {
   integer n;
   
  
   
   for (n=0; n<6; ++n)
   {
       RotFromEdge(45, Z_NEG, X_AXIS);
       llSleep(1.0);
       RotFromEdge(-45, Z_NEG, X_AXIS);
       llSleep(1.0);
   }
             

 }
}

I have tried this sript and it works fine except when attached. I have added passing in the local rot of the root (I am rotating a child) and tried that but can not get it to work.

I am making a locket on a necklace that opens - all is fine when not attached.

When attached I get weird rotations:

Here is my change in the attached rotation test:

if(llGetAttached() != 0)
{//--need a new function for this, with the roots rotation passed to it
rotation R = llGetRootRotation();
llSetPrimitiveParams([PRIM_POSITION, (llGetLocalPos() + rotOriginBefore - rotOriginAfter)/root_rot, PRIM_ROTATION, newLocalRot/R]);
return;
}

Could someone post the wrokaround for attached rotation of a child prim?

TIA,

Frank

heyas frank!

okay, the problem is that on an attachment, llGetRotRotation gets the AVATAR's rotation, not the root prim, which is what we need.

what i usually do for attachment children is on rez, have the root prim link message its rotation to the children. (change the rotation to a string and pass it via the string section of the message.)

then in the child i want to rotate, in the link message event, i decode the rotation string and set it to my RR (or your R in this case).

then just use that variable when you want to rotate the child.


ps: i have no idea what would happen if the user changes the rotation of the attachment while wearing it. probably blows it up again. ::sigh::

In that scenario you might be able to catch the rotation with the moving_end() event handler and update the variable which stores the root rotation via linked message.

doesn't moving_end (and start) also happen when the avatar just walks around, though? does it happen if an attachment is on your arm and the avatar waves the arm?

i know some attachments 'sense' motion/speed of an avatar. or themselves ON an avatar.

wouldn't it be like changed and.... something? :/ i haven't run into it as a problem yet, that's why i dont know ;D

You can not use the changed() event for this as far as I know.

So, does this thing work or not?
And if it works, which of the multitude of snippets of scripts works?
/me shakes her head in self-pity

Good question...did you modify the WIKI to add this important info? You can, you know.

the only wiki i use is rpg stats, and rpgstats says not to make edits to the wiki, because it's going to re-download from somewhere else.

so, no.


ive seen "if(changed & change_somekindorother)" but only in passing. as you can tell, im not exactly sure how/where this goes. and i still haven't tested any of this. i know, i get distracted by -- ooh, lookit that neat avatar!


dr fran, the LAST and LATEST script is the one that works, of course. ;) gotta get the latest version.

Hi.
I just opened a thred to discuss methods how to optain the local rotation of the root object:
/54/07/210449/1.html#post1676079