This doesn't make sense to me - but perhaps it does to you

[Phil Deakins]

I'm writing a root prim script in which moves a child prim. Things weren't working as I thought they should, so I did the following test:-

The root script got the PRIM_POS_LOCAL of the child prim, so the child's position is with respect to the root, an d immediately set the child prim to that PRIM_POS_LOCAL.

To my way of thinking, the child prim shouldn't move, but not only does it move to the opposite side of the root prim, but it also sinks 0.18 into the floor.

Why would the child prim's position change so drastically when setting it's position to the exact vector value that the script has just obtained as its position? I'm flummoxed.

[steph Arnott]

I can not read your script snipet so have a guess. The local is relative to the root, so going by what you wrote seems the root prim is rotated, also the local needs to set with

llSetLinkPrimitiveParams

 this should work.

llSetLinkPrimitiveParamsFast(child_prim_linknum,[PRIM_POS_LOCAL,<x,y,z>]);

 

[Rolig Loon]

---

steph Arnott wrote:

I can not read your script snipet so have a guess. The local is relative to the root, so going by what you wrote seems the root prim is rotated, also the local needs to set with

llSetLinkPrimitiveParams

 this should work.

llSetLinkPrimitiveParamsFast(child_prim_linknum,[PRIM_POS_LOCAL,<x,y,z>]);

 

---

Well,

llSetLinkPrimitiveParamsFast(child_prim_linknum,[PRIM_POS_LOCAL,<x,y,z>*llGetRot()])

After all, as you pointed out, correctly, Phil forgot to take rotation of the root prim into account.:smileywink:

EDIT: I'd say Ooops quietly and delete my post, but that would be intellectually dishonest.  I screwed up.  :smileyembarrassed:

[steph Arnott]

oops, forgot the rotation.

[Qie Niangao]

Wait, I'm confused. Isn't PRIM_POS_LOCAL already in the root prim's coordinate system?

If so, might something else be wrong with this script?

[Ron Khondji]

If I do this:

vector lPos;integer S;default {    touch_start(integer total_number) {        if (S = !S) {            lPos = llList2Vector(llGetLinkPrimitiveParams(2, [PRIM_POS_LOCAL]), 0);            llOwnerSay("POS_lOCAL vector = " + (string)lPos);        }        else {            llSetLinkPrimitiveParams(2, [PRIM_POS_LOCAL, lPos]);            llOwnerSay("POS_LOCAL set");        }    }}

the child prim does not move, no matter what rotation the object has.. 

So you must be doing something wrong.

[Rolig Loon]

Right you are.  Sorry.  Fur on the brain this morning.

[steph Arnott]

Assumed there was, but as i was makeing a guess just thought be easy to be basic and let them use that info.

[Phil Deakins]

Panic over. I was about to post the relevant lines of code when I saw that I was getting the local position of one child prim (i) and applying it to another (i+1). Sometimes it's embarrssing :manembarrassed:

The words 'wood', 'trees' and 'tired brain' spring readily to mind

[Rolig Loon]

Ah, well, there's a bit of that to share today.  :smileyembarrassed::smileytongue: