Periodic Timed Event VS While Loop

[jleips2600]

Is a Period Timed Event every .03 seconds any different (better, worse, the same?) as a While Loop running for a significant amount of time with a .03 second wait?

 

[Weep]

The timer will actually run at 0.0625s intervals, and so will the wait. However, a 0.0s wait will let the loop run at 0.03125s intervals. [source]

Consider using Smooth Timers?

 

[Arkless]

Is a Period Timed Event every .03 seconds any different (better, worse, the same?) as a While Loop running for a significant amount of time with a .03 second wait?

a timer event fires a trigger
i think a fired trigger creates a new thread

Wait + loop only stops/restarts the running trigger and I believe it's faster. (no testing done thought)

 

[Zwiebelchen]

Waits also create new threats. I think it doesn't make much difference in terms of efficiency, so use the method that you think is more comfortable.

 

[Sensang]

Another difference is that you can refer to the "triggering unit" within a while loop. Something you can't do within periodic events.

 

[jleips2600]

But running a while loop with a .0625s wait for 20 seconds here and there would be a lot better than a constant periodic event every .0625s I assume?

 

[Zwiebelchen]

But running a while loop with a .0625s wait for 20 seconds here and there would be a lot better than a constant periodic event every .0625s I assume?

No, why should it?

 

[Arkless]

No, why should it?

[del]To end this conversation I just tested it >.< we got that awesome trigger debugger anyways. Made one periodic event and one loop and stopped after the same time:

Trigger1:
Events
Timer - Every 0.0625 seconds of Game Time
Local Variables
Conditions
Actions

Trigger2:
Events
Timer - Elapsed time is 0.0 Game Time seconds
Local Variables
Conditions
Actions
General - Repeat (Actions) forever
Actions
General - Wait 0.0625 Game Time seconds


stats:
Trigger1:
!!! Total Time (ms) 41.127
Trigger2:
!!! Total Time (ms) 3.334

We can see, that the "repeat forever" choice is WAY faster. Anyways, it was running for 3:04 so the time consumed is very low.

We should note as well that it always displays that Trigger1 has a thread waiting for 0.0625 gametime while Trigger2 isn't listed at all.


My conclusion (there is always the possibility of a mistake somewhere): "repeat forever" clearly beats a periodic timer events and doesnt delete my locals [/del]

It seems like the the debug window has some bugs <.< From time to time the runtime of the repeated trigger becomes a negative value.
After a bit more testing I found out that the repeat forever nees something between 1.5 and 1.7 times the calculating time to execute.

 

[Ancanus]

[del]To end this conversation I just tested it >.< we got that awesome trigger debugger anyways. Made one periodic event and one loop and stopped after the same time:

Trigger1:
Events
Timer - Every 0.0625 seconds of Game Time
Local Variables
Conditions
Actions

Trigger2:
Events
Timer - Elapsed time is 0.0 Game Time seconds
Local Variables
Conditions
Actions
General - Repeat (Actions) forever
Actions
General - Wait 0.0625 Game Time seconds


stats:
Trigger1:
!!! Total Time (ms) 41.127
Trigger2:
!!! Total Time (ms) 3.334

We can see, that the "repeat forever" choice is WAY faster. Anyways, it was running for 3:04 so the time consumed is very low.

We should note as well that it always displays that Trigger1 has a thread waiting for 0.0625 gametime while Trigger2 isn't listed at all.


My conclusion (there is always the possibility of a mistake somewhere): "repeat forever" clearly beats a periodic timer events and doesnt delete my locals [/del]

It seems like the the debug window has some bugs <.< From time to time the runtime of the repeated trigger becomes a negative value.
After a bit more testing I found out that the repeat forever nees something between 1.5 and 1.7 times the calculating time to execute.

I've attached a map that demonstrates they have the same effect. Run map with debugger and look at the variables PeriodicTimer and RepeatForever.

 

[Arkless]

However, try running them independently.
(Check your Threads tab when running them both at the same time.)

I've attached a map that demonstrates they have the same effect. Run map with debugger and look at the variables PeriodicTimer and RepeatForever. (Show Constants)

Duh, i probably don't get your point, but from my view you totally failed to answere to the given question... Question being: "Which way to repeat actions needs less runtime?"

 

[Ancanus]

Duh, i probably don't get your point, but from my view you totally failed to answere to the given question... Question being: "Which way to repeat actions needs less runtime?"

That's alright, I was elaborating on your research (which was an excellent initiative, by the way) not on the question. As for the question, both evaluate the same amount of times, on a given time.

 

[Arkless]

That's alright, I was elaborating on your research (which was an excellent initiative, by the way) not on the question. As for the question, both evaluate the same amount of times, on a given time.

Yea, that's still not the question <.< We know that both will execute the same amount of times within the same time. We already got that in the smooth timer discussion thread. The question at hand is how long the computer needs to execute what we want (reason is lag).

If for example one event from the periodic timer event needs 0.01 seconds (that way to high, just for the example) we wouldnt be able to use more then 6 0.0625 periodic events (or it would perma lag).

If we only wanted to repeat an action we could also use repeat forever, so we do the same test and get 0.015 (for example). Which would mean we can only use that 4 times with an 0.0625 wait.

 

[Ancanus]

Yea, that's still not the question <.< We know that both will execute the same amount of times within the same time. We already got that in the smooth timer discussion thread. The question at hand is how long the computer needs to execute what we want (reason is lag).

I'm not sure if I'm following, but if both execute the same amount of times in the same time, does that not imply that the computer spends the same amount of time in both?

Edit: Oh! I think I see what you mean. Are you talking about the evaluation time of the actual events? If so, it would make sense that repeat is faster, since it just calls a sleep function. Whereas, the periodic event is instantiating an implicit timer.
However, note that in the map I posted, it is the repeat forever trigger that takes more time (which conflicts with your observations). This could be because it counts the wait time action as part of the trigger running, whereas the periodic one doesn't. This leads me to believe they both "need same amount of runtime", or at the very least, you can't use these numbers to draw a conclusion upon.

 

[Arkless]

This could be because it counts the wait time action as part of the trigger running

Not at all <.< the cound in milliseconds only counts the time needed for executing actions. My problem is that the debut window somehow jumps to negative values after some time (got no idea why). But for the first few seconds it runs smoothly and shows a difference between 1.5 and 1.7 (you might have noticed that I [del]striked[/del] my previous post after noticing). The [del]periodic[/del] loop seems to be 1.5 to 1.7 times slower.
edit: typo, sry >.< i ment loop, man, i gotta take more time for my posts >.<

 

[Ancanus]

You are right about the wait time not affecting the run time, I just tried it out.

Edit: I added a few more tests. From faster to slower it seems to be:

1. RepeatForever
2. CommonTimer
3. PeriodicTimer
4. RerunTrigger