need help detecting an avatar with a phantom prim

[RhaDo2496]

hoi

is it possible to use a phantom prim to detect an Avatar standing inside it?

[Rolig Loon]

I'm not sure what you are asking.  You can detect collisions with a phantom object if you script the object to recognize the collisions with llVolumeDetect(TRUE).  And if course you can put a llSensor in anything, whether it is phantom or not.  

[Rachel1206]

No, as you noticed a phantom object will not  detect collision etc, so only by using for example llSensorRepeat()

Set the radius equal to the box in question and keep the timer with a high value like whole seconds, the SIM impact will be low. Just remember sensors use resources and have an impact on SIM performance.

An elegant solution would be to link two objects, where the prim object is default prim say 1 x 1 x 0.001 meters and invisible,  link the second object say 1 x 1 x 2.5 meters and set physic shape to none in the child object. Just use collision_start in the root prim and hence avoid llSensorRepeat etc.

 

 

Edited April 10, 2018 by Rachel1206

[Rolig Loon]

31 minutes ago, Rachel1206 said:

No, as you noticed a phantom object will not  detect collision etc

Not true.  As I noted earlier, if the OP wants to detect collisions, all he has to do is include llVolumeDetect(TRUE) in the state_entry event.  I just wasn't clear about exactly what the OP wants to do.

[Qie Niangao]

30 minutes ago, Rachel1206 said:

An elegant solution would be to link two objects, where the prim object is default prim say 1 x 1 x 0.001 meters and invisible,  link the second object say 1 x 1 x 2.5 meters and set physic shape to none in the child object. Just use collision_start in the root prim and hence avoid llSensorRepeat etc.

Not sure what's intended here, but I don't think this will do what's expected. A linkset element with "None" physics type will never detect collisions. Maybe I'm not understanding its intended purpose here. But anyway, Rolig's llVolumeDetect(TRUE) approach is the standard way to avoid sensors (or the other approach, monitoring avatar locations in llGetAgentList()), and sure seems to be what the OP asked.

[Rachel1206]

1 minute ago, Qie Niangao said:

Not sure what's intended here, but I don't think this will do what's expected. A linkset element with "None" physics type will never detect collisions. Maybe I'm not understanding its intended purpose here. But anyway, Rolig's llVolumeDetect(TRUE) approach is the standard way to avoid sensors (or the other approach, monitoring avatar locations in llGetAgentList()), and sure seems to be what the OP asked.

Root prim set to prim - linked object set to none - you can drive a passenger train through it and it will detect it

[Qie Niangao]

I must be misunderstanding. If the train collides with the root prim then a collision event is generated, but not if it only passes through the "None" child prim. So... maybe that child prim is merely intended to be decorative?

[Rachel1206]

24 minutes ago, Qie Niangao said:

I must be misunderstanding. If the train collides with the root prim then a collision event is generated, but not if it only passes through the "None" child prim. So... maybe that child prim is merely intended to be decorative?

OP asked when an avatar was standing inside the cube in question. and of course it will only work if an avatar, object - whole planet - get in touch with the root prim.

Now the usage of this kind of solution could be many, say we put on a semi-transparent texture on the child cube, for example:

• teleport platform - stand by/in use
• old woman crossing road, where it changes to screaming old woman when we hit
• interactive visitor greeter
• signal at intersection

And yes to llVolumeDetect() as the standard way, I was just thinking out of the box on other approaches.

Edited April 10, 2018 by Rachel1206