User talk:Chibbchichi Resident

llGetAgentInfo Example

I saw your recent contribution to llGetAgentInfo, but I felt it had some problems that needed addressing, so I took the liberty of moving it to here. Your attach event resets the script, forcing it back through state_entry. Within state_entry you assign the owner's key to the string 'id' (strictly should be a key variable). But then you attempt to test if the object is being detached by checking if id is NULL_KEY - but it can't be. That test needs to be done inside the attach() event. Stopping the timer from within state_entry is pointless, as no timer will be running first time nor on script reset. Within timer(), the information you gain from llGetAgentInfo() is not only about whether the agent is typing. It could for example include mouselook indication, or if they are flying. So testing this data for a change in typing status is not totally valid. You must isolate the AGENT_TYPING bit straightaway. I won't try and fix the whole script, but perhaps this would work for the timer code. The rest needs a bit of a rethink. Omei Qunhua 11:18, 26 December 2013 (PST)

<lsl> timer() {

   integer TypingNow = llGetAgentInfo( llGetOwner() ) & AGENT_TYPING;
   if (TypingNow != gLastAnimation)
   {
        if (TypingNow)
           doStuff();
        else
           stopStuff();
       gLastAnimation = TypingNow;
   }  

} </lsl>

[V3] <lsl> // A simple typing override example. // Modified the example-LSL of llGetAnimation to make it work with llGetAgentInfo. // Also included an easy way to hide and show the whole object depending on current typing state. // For example: Show a keyboard under avatars hands when avatar types.

key gOwner; // the wearer's key integer gWasTyping; // last state of typing seen

// User functions Initialize(key id){

   // in case the script was reset while already attached
   if (llGetAttached() != 0) {
       if (id == NULL_KEY) { // detaching
           llSetTimerEvent(0.0); // stop the timer
       }
       else { // attached, or reset while worn
           llRequestPermissions(id, PERMISSION_TRIGGER_ANIMATION);
           gOwner = id;
       }
   }

}

// Event handlers default {

   state_entry() {
       Initialize(llGetOwner());
   }

   attach(key id) {
       Initialize(id);
   }

   run_time_permissions(integer perm) {
       if (perm & PERMISSION_TRIGGER_ANIMATION) {
           llSetTimerEvent(0.25); // start polling
       }
   }

   timer() {
       integer typingNow = llGetAgentInfo(llGetOwner());

       if (gWasTyping != (typingNow & AGENT_TYPING)) { // any change?
           if(typingNow & AGENT_TYPING){ // set whole object visible when typing
               llSetLinkAlpha(LINK_SET, 1.0, ALL_SIDES);
           }
           else{ // set whole object invisible if not typing
               llSetLinkAlpha(LINK_SET, 0.0, ALL_SIDES);
           }
           gWasTyping = typingNow; // store away for  next time
       }
   }

} </lsl>

I agree with Omei here. The purpose of the examples in the wiki pages is to be educative about what the function does. This script is, at best, an example of how to put a bunch of functions and events together in order to make a typing overrider. Not really educative on what llGetAgentInfo does by itself. The individual functions are not the correct place for putting a script that requires mixing so many functions and events. The example that is already present there is enough, and serves as a good illustration of how an educational example looks like. --Pedro Oval 15:07, 29 December 2013 (PST)