llSensor

I have some rezzer scripts that are now working in a new and exciting way for no obvious reason.

Basically, I put a script into an object, and that script runs an llSensor for another object, so it can determine its relative position.

The script uses the ACTIVE constant in the function arguments, as this catches physical or active scripted objects as the sensor target. It's worked like this for over two years.

I passed the scripts onto someone and he has been struggling to get them to work.

I did some experimenting, and it seems that something has changed. The target object has a listen event in it- this has previously qualified it as "active scripted" and so should be caught by an llSensor using either ACTIVE or SCRIPTED arguments.

llSensor will only find the object if the argument is changed to state PASSIVE. This is only on newly compiled versions of the script- my old scripts I just drop into prims still work fine, but if I create a new object with a simple llSensor script, then neither SCRIPTED nor ACTIVE arguments will find the target object- even when the target object has a listen in it.

Did I miss some change, is it some kind of lag/timeout thing? Has anyone else had this happen? My current dirty fix is to add a second llSensor line with PASSIVE as an argument, to catch all object types.

Wah!

I haven't noticed this, as I mostly scan for people, not objects. But it does sound like something has changed. You should probably start a JIRA on it, I haven't seen anything to indicate this was changed on purpose.

I have modified the scripts you originally gave me, so maybe it is not the best test, but the "Rezzed Object" script is nearly the same. The "Rezzer" script is different in that I have one Rezzer box for each build instead of your original method.

Anyway, I tried your experiment by setting up a new Rezzer box with old scripts for the the Rezzer and the Rezzed Objects, and it worked normally. Then, I created new scripts and dropped them into the Rezzer Box and each object of the build, then took the objects of the build into inventory, dropped them in the Rezzer box and did not have any problems.

Maybe I didn't read your process to duplicate the problem right. I can send you the scripts if you would like.

My hair has grown back since I had this problem (and not only me, but someone who was struggling with the scripts).

It seeeeems to have been something to do with especially crappy SL performance- it was one of those nights I think when LL was sending out blue boxes telling people not to do anything.

It's wierd, though because copies of my existing rezzers worked fine- these still involved a "fresh" llSensor in their on_rez which must have had to find the rezzer box as the source for their positions. But newly written llSensors had to scan for passive, not active objects to find their source rezzer box.

The solution seemed to be "belts and braces" by doing two sensors, once for passive, once for scripted. One of them should work, whatever the SL weather