detect UUID

I manage to get a gadget that will detect UUID on avatar, but it can't detect all the avatar's UUID. ANybody know why? And how can I create a gadget that can detect all avatar UUID? Any tutorial or sample to learn from?

depends on how it's detecting it for why it might be failing....

if you can get withing 96m of them it's pretty much guaranteed you can get their key... a simple sensor and dump the name and key that's detected (might need to be closer if there's more than 16 people around)

no tutorial, but the following portal pages will help
http://wiki.secondlife.com/wiki/LlDetectedName
http://wiki.secondlife.com/wiki/LlDetectedKey
http://wiki.secondlife.com/wiki/Sensor

Do you mean UUID's of attachments? As far as I know, you can't specifically get those attachments UUIDs, the best you can do is just detect if they are wearing scripted ones with a sensor.

Void is talking about the avatar's uuid, not the id of attachments, you cannot detect those as far as I know, and even so, why would you want to know it?

I assumed OP was talking about actual AV keys (which look the same as an object UUID) ... to my knowledge there is no way to grab the UUID of ANY attachment unless it sends out a communication that can be picked up (like say, chat's a message openly)...

past that I don't even know if a custom client could do that. but LSL won't.

I took it as meaning the OP has encountered the limit of 16 reported results for llSensor and is looking for a work-round that will report the uuids of everyone in scan range, which I am afraid I don't know how to do.

that makes sense too... one of the ways is to limit the arc to less than a full sphere, and rotate the sensor to cover other areas... or use multiple sensors in that config....

the rotations for a tetrahedal (4 sided) sensor net are

Tetrahedron Face Rotations
< 0.000000, 0.000000, 0.000000, 1.000000>
< 0.577350, -0.577350, 0.000000, 0.577350>
< 0.000000, 0.577350, -0.577350, 0.577350>
<-0.577350, 0.000000, 0.577350, 0.577350>

with an sensor angle equal to the PI - the dihedral angle IIRC... ~111deg if I've got that right

I have rotations for the 8 and 12 if anyone wants them, also... 6 should be say to figure out, and 2 is obvious. 20 can be gotten from the twelve by figuring out the centers of each set of three touching faces.

Interesting. What do you use for the 12? Dodecahedron or edge-centred cube (cubic close packed, easier to calculate)? In all of these, is the sensor scope a cone? and does that mean there is always either overlap or gaps? (except the 2 sensor case!).

not through lsl, but on Emerald i think if you select someones attachment with the edit window already open it will let you get the uuid like any other object (the button at the bottom right corner of the first tab

I used the vertex points from icosahedron which are easy to calculate, which are incidentally the face directions for the dodecahedron (and vice versa). then I use rot between on the generated points.... you can do the same to get the octohedron faces from a cubes corners. if you want to orient any face forward you can apply a rotation to the point before hand, or after conversion to a rot.


yes cone, yes minimal overlaps ,no gaps with the provided formula of PI - dihedral angle in radians. the dodecahedron provides the least per edge overlap, but I'm not sure if it has the least overall (twin hemispheres obviously wins there on both counts.)

ETA:
the truncated icosahedron (AKA the soccerball) and any of the other truncated platonic solids are possible but they'll use two different cone sizes so are harder to calculate/use properly.