What does llGetRot() return when called in a script that is in a non-root prim of an attachment?
Ratany Resident 05:13, 19 February 2014 (PST)
- It follows the same strange logic that applies to llGetPos, that is, you'll only get reliable results if it's attached to ATTACH_AVATAR_CENTER and the attachment's root is ZERO_ROTATION relative to the avatar. This is all because the simulator really doesn't know anything about the shape of the avatar's skeleton, plus animations are a viewer effect. I'll copy the note over to here. --Cerise Resident 05:33, 19 February 2014 (PST)
did this change when they reworked/added attachment points? it used to be that attachment had their own "global" space, equivalent to, but separate from, region global. Meaning that a rotation around a given attachment point would report the same, even if an object the avatar was seated on was changed in rotation (so not just animation), as it would only be relative to the attachment point.
-- Void (talk|contribs) 09:38, 19 February 2014 (PST)
- I did check this to make sure it was the current behavior. This stuff was ridiculously hard to explain clearly before we had that avatar center attach point as a reference. It's not so much that there is a local reference, as the simulator doesn't know any better and assumes all the attachment roots are lined up with the avatar. --Cerise Resident 12:35, 19 February 2014 (PST)
- rotation#Single_or_Root_Prims_vs_Linked_Prims_vs_Attachments tries to explain it. -- Strife (talk|contribs) 13:26, 19 February 2014 (PST)