Talk:LlSetTimerEvent

It would be really great if this Wiki contained information that lots of other scripting wikis contained such as whether a timer, once set, persists across state changes.

(Ramana Sweetwater 2009/01) LlSetTimerEvent will trigger timer events in all states within the same script where the request was made.

• If there are multiple states with timer events in the same script, always use a conditional to determine which code should be executed in the timer event.

---

There is a somewhat quirky behaviour across state changes: Say you run a timer at 10-second interval, and it fires at time 0, 10, 20 etc... If you switch into a state without a timer event handler, the timer still "ticks", but of course, nothing is triggered. If you return to the state with a timer event after less than 10 seconds, the next event will fire "on schedule" at 30, 40 etc... But if you stay in the second state longer than a timer cycle (say, 15 seconds), the timer event *will immediately fire* upon return, and then pick up a 10-second interval *after that*, i.e. 35, 45, 55 etc. -In fact, it tends to return to a schedule looking like 35, 44, 54, 04, i.e. the first interval being a little shorter, but that is likely just lost precision due to the processing during a state change.

The behaviour is easy to observe with a script like<lsl>default

{
    state_entry()
    {
        llSetTimerEvent(10.0);
        state one;
    }   
} 
state one
{
    touch_start(integer total_number)
    {
        state two;
    }    
    timer()
    {
        llOwnerSay((string)llGetUnixTime());
    }
}
state two
{
   touch_start(integer total_number)
    {
        state one;
    }
}</lsl>Touch it to go into state two, wait for 15 seconds, and touch it again. Compare to what happens if you only wait 5 seconds.

Tali Rosca 21:23, 15 November 2009 (UTC)

---

IMPORTANT ??? Exactly WHAT is so IMPORTANT about the IMPORTANT notes on the llSetTimer() event page? Especially what is so important and one individual's (horrible) personal coding style? And that note on llFrand() producing zero ... is this really the right place for it?

And anyway the same editor has produced his 16th non-compilable example script here (undeclared variable 'gap'). So I take it that gives me Carte Blanche to try and return this page to some sanity. Omei Qunhua 12:52, 30 December 2012 (PST)