Yes I do have activation and deactivating groups.
That "could" be it, but I doubt it.
My gnaw has under 75 objects in frame at a time though, and only one of those objects is bouncy ball.
It's odd that this is happening since I've played games from the app store that are so much more intense with no slow downs at all.
"Always set x position to x position + 2"
It should be "+ 2 * elapsedtime".
I'm still expirencing performance issues on the iPhone 3GS
It starts out fast enough, but then about 2 seconds in cuts to about half the frame rate.
On the iPhone 4 it runs full speed throughout. I'll have to do some more testing to see if the jumping ball glitch is still there though.
I have debugged this to death and it doesn't seem to be a fault of the runtime itself (CPU or GPU usage). It could be a timer issue though. I'm looking into this.
If the runtime isn't at fault what's the problem?
It can't be the iDevice since it runs a lot more complicated stuff than this at the same speed.
This game is literally just blocks on a black background with only 3 moving objects at most.
I don't think it's my code. It's very basic code, the only complicated part is the level loading, but that bit doesn't run during the gameplay so it shouldn't affect framerate.
I made it since I figured it would be the least likely to have these kinds of issues. Something quick and easy, but it's not turning out that way.
So at this point should I consider dropping the resolution of the game back to the iPhone 3GS resolution and just forget about retina display? I would also have to make it iPhone only instead of universal since it wouldn't look as good scaled on the iPad.
Thanks for looking into it!
OK with more testing the timing issues occurs on the iPhone 4, the iPhone 3GS, and the iPad 2 it happens less often on the iPhone 4 and iPad though. So the faster the processor the less often it happens.
This issue affects all built-in movements, and doesn't affect anything that's custom made. So the ball, the laser, and the powerups that are dropping all skip at the same time.
This also affects collisions every now and then, sometime it triggers the ball to collide with all the blocks at once, so all the blocks on the screen get destroyed sometimes, except for the multi-hit blocks which remain, so this let's me know that this is an issue with the collision test when the skipping occurs.
I noticed that the skipping happens more often towards the beginning of gameplay then starts happening less often towards the end. It might just be chance though, so I'll have to test more to confirm this.
Have you noticed the same behavior in your tests Andos?
This seems to be something similar that the delta-time issue with the path movement (now fixed).
UrbanMonk:
Francois fixed the bug where it destroys all bricks. It should be in RC2 AFAIK.
I experience different behaviour with the movement on my 3GS. I don't have any other devices I can compare with sadly.
I'm still doing some profiling to figure out exactly when it happens and see if I can find a pattern.
I updated to the latest build already and tested. You seem to be right. I can't get the brick destroying bug to happen again. I just assumed it was still there since it happened randomly, but rarely.
But all the other bugs I mentioned are still there.
I'm glad you're on this!
I'm happy to report that the game runs perfectly in all the other (non-beta) runtimes, including java mobile, it runs at a full 30 fps at the lower resolution on my blackberry curve, torch, and storm.
Posting Permissions
