Why no Agent name on sensor detect?

Hey all,

In this code fragment a sensor event is raised and I call llDetectedKey(0) and get a valid key for an agent (me for example).

But when I call llKey2Name() I get an empty string. Was up?


key agentKey;
string agentName;

sensor(integer total_number)
{
for ( i = 0; i < total_number; i++ )
{
agentKey = llDetectedKey(i);
agentName == llKey2Name(agentKey);

llSay(0, "Agent - UID: " + (string) agentKey + " Name: " + agentName);

Output from llSay() shows a proper key but no name. I have used llKey2Name in other cases (not during a sensor event) and it has worked. But not here.

Anyone know why? TIA.

For llKey2Name to work the agent must be in the same sim.

Try llDetectedName(i);

 agentName == llKey2Name(agentKey);

needs to be

 agentName = llKey2Name(agentKey);

Single = for assignment, double == for comparison.

And you can use llDetectedName(i) directly, you don't need to go through the two steps of llDetectedKey followed by llKey2Name.

llKey2Name only works for avies in the same sim, so will get nothing if not in same sim.

Use Dataserver for that information if online/offline same sim, does not matter..

http://secondlife.com/badgeo/wakka.php?wakka=llRequestAgentData

DUH! Of course I copied that right out of a prim and in fact I had the double equal as shown in this thread. Double DUH! Never noticed it. Stared at it for 10 minutes.

In this case I need the agent's name and key (both) otherwise I would just call llDetectedName(), but I could see how you might assume I was a bleeding idiot after the double equal screw up! Ha ha.

I'm not quite getting that - you can get the name with llDetectedName, and the key with llDetectedKey, what's the problem with using both if you want both?

No problem, per se. But intuitively it seems like doing a lookup on the key to get the name should be less expensive then calling llDetectedName. Presumably LL refers internally to avatars as agents using the key. I suspect the body of llDetectedName contains a call to llDetectedKey and then a call to llKey2Name - that would make sense and thus why incur an additional call to llDetectedKey considering I already have the key. That is my thinking. Perhaps I am completely wrong about the implementation of llDetectedName but there you have it.

Not necessarily. I think internally, the llDetected functions refer to an array (which the integer indexes into) of structs of data (name, key, position, rotation, &c) that store information about the detected stuff. (I'm probably wrong.) What llDetectedName offers you is the capability to retreve the name of an object/avatar that is currently (or has moved) outside the simulator of the sensing object. (Sensors occasionally pick up things outside the sim, when their range extends over a sim border.)
==Chris