One-shot Trigger Disabling

[Hooman]

Heh, almost forgot I was supposed to post this info.

If you have a trigger that is set to only fire once, then it will disable itself after it fires. The question I got from Hidiot, was at what point does the trigger get disabled.


Here's the sequence of events.

Code: [Select]

1) For each trigger:
 Â   2) Check if the trigger should fire
 Â   3) If the trigger fires, then get the callback address (the function you export from your DLL)
 Â       3b) When getting the callback address, call Disable() if it's a One-shot trigger
4) For each trigger that fired:
 Â   5) Call the trigger callback

This is important because, by the time your callback function executes, Disable() has already been called on that trigger. This means you can re-enable it in the callback. If it got disabled after the callback, then it would have forgotten any changes you tried to make.

This also means you can re-enable other One-shot triggers that are firing this tick, but haven't yet had their callback functions called. That is, if you have two triggers, A and B that will both fire on the current tick, then the callback for trigger A can re-enable trigger B, even before the callback for trigger B executes. As you can see in the above sequence of steps, all the triggers are checked and disabled if needed, before any of the callbacks are called.


Edit: I'd also like to point out that Disable() is not the same as Destroy(). If you Disable on a trigger, it still exists, and takes up one of the 255 ScStub objects, but won't fire. If you call destroy, then the space allocated for it is freed and can be reused.
 

[Eddy-B]

Biggest question on people's minds right now would be:
If i destroy the one-time trigger from inside the callback function to save resources (read: ScStub objects), will that work or do i get some kind of exception?

[Hidiot]

It will work.

I've seen it work.

[Hooman]

From what I've read of the code, I see no reason why that wouldn't work.

Of course I've never tried it.