LSL functions you should NEVER call

[Graham Mills]

Not sure whether it will help here but I saw from HGBiz that there's a move to implement the Phlox script engine used in Inwz.

I noticed in the thread there that there were some adverse comments about llSetTimerEvent as well as llSleep so I thought I would have a look albeit on a limited scale. Firstly, however, I had a look at llSleep. I don't know how useful it is as an observation but my experience with 24 scripts running was that llSleep was OK up to 30 sec but no good thereafter (subsequent commands were not executed). This was true even for a single script so far as a 50 sec sleep was concerned. However, even then the 24 llSleep scripts (mostly 30 sec but including one incremented to 50 sec) failed to interfere with llSetTimerEvent running separately for 70 sec in a further 12 scripts. I didn't confirm the accuracy of the intervals but I suspect it wasn't entirely as one might naively have expected.

I'm not saying that I'd recommend using llSleep but I'd certainly avoid it for intervals longer than 30 sec. Incidentally, it may be that the commenter was referring to 0.9 -- it wasn't clear.

[Handy Low]

The problem with long sleeps is that they can make the event they're in exceed the maximum time allowed for an event. In that case, the event is killed and, as you say, subsequent commands are not executed. This would also happen if, for example, you had a loop that took a long time to execute - it's not a problem with the llSleep() function per se.

I'm not actually sure what the value for the maximum event time is in Kitely, but 30 seconds seems about right.

[Alexina Proctor]

Is llSleep() still a problem in the current OS version?

[Alexina Proctor]

Yes Alexina, it is on Xengine (which is what Kitely uses).

[Christine Nyn]

The main problem with using llSleep is that it quite literally sleeps. It suspends all operation on that script thread in the simulator but it doesn't free the thread up for other scripts to use as llSetTimerEvent would do. Get a few of those on a sim and the whole thing will grind to a halt as other scripts are denied access to resources that allow them to function. If, as in the OP, the "sleep" is 30 seconds or longer the script will be stopped in any case as it exceeds the standard OS event time-out limit.