The Super-Ball Effect

[JagdPanther101 0]

Has anyone else noticed that rocks tend to act like giant slingshots (like Pinball slingshots; when the ball contacts them, they quite forcefully smash the ball away) when a vehicle contacts them?

I see it happen in 1x Time Mode a lot. I'll run into a rock accidentally and my vehicle (usually the M113 right now since all I'm doing is unlocking items in the Armory) get thrown into the air like it were a super-bouncy-ball. In the 2x and 4x time mode (hey, I wanna gain points fast) it seems to actually be multiplied several times. It's only time moving faster, not my vehicle, yet my vehicle seems to act like a feather in the wind when it hits the rock.

Does anyone else have this problem? It's rather annoying.

[-Puma- 2]

yes it's collision bug, when vehicle collides in certain objects. They've tried to fix this many times.. (havent seen it in 1.07b tough) BIS has some elevation problems (maybe suma or maruk should take a math lesson or something )

5km is 4995 meters higher than 5m...

[4 IN 1 0]

by making the game even more laggy?

[-Puma- 2]

by making the game even more laggy?

what u mean by that??? i was only pointing out that bis has obiously put a incorrect value when vehicle hits rock, Instead of 5km jump value, there should be 5m. Same goes to the gunner exiting humvee in 5m in the air, where they shuld land 0m in the ground.

So basic math for BIS

[frederf 0]

It's in the 1.07 changelog.

[JagdPanther101 0]

It's in the 1.07 changelog.

Ah, that's fabulous. Thanks for telling me.

[Col.Flanders 0]

I noticed something else with the rocks. I suspect it has to do with terrain detail but I noticed that when shooting very close past rocks, you often hit an invisible "something" (probably reserved for the rock geometry on higher terrain detail). The cool thing was that when this happened, I noticed a very good ricochet. The bullet deviated from where I was shooting through nothing but air and I saw the tracer deviate about 30 degrees and it left a puff of dust in the ground (way off target but I was impressed.)

[Vassago 0]

They need to completely scrap their awful physics system and go with something from Havok or PhysX. It's been what, 7years or so? The engine needs a better overhaul.

[fireship4 0]

I think ragdoll was attempted but it was too laggy. I wouldnt mind seeing them release it as an addon for those who like playing on smalll maps/who have better systems though...

[Average Joe 0]

No Chance of that happening, ragdolls gotta be hardcoded as far as I know.

[Metal Heart 0]

Quote[/b] ]i was only pointing out that bis has obiously put a incorrect value when vehicle hits rock, Instead of 5km jump value, there should be 5m

...

So basic math for BIS

The only obvious thing is that you have exactly zero understanding of basic physics programming If it were only a matter of a single incorrect value the problem would be consistent.

I think there is a fundamental design flaw in the engine of not having the physics and collisions independant of fps. That is very evident when comparing collisions on 30 fps to collisions on 5 fps for example, low fps causes problems such as the pinball mode while the exact same collision on high fps would always results in a quick halt with no air time.

[Vassago 0]

Havok doesn't instantly mean 'ragdoll'. Havok physics are used on vehicles, objects and flight models as well.

And metal heart, that's not entirely right. I hit a small rock with the vulcan last night, and I was getting 40-50fps. I flew at least 30ft into the air, damn near flipping. There's something severely wrong at the core level of the physics engine. And it needs to be fixed.

Does anyone have VBS/VBS2? If so, are these same medicore physics flaws still there?

[weegee_101 0]

I think there is a fundamental design flaw in the engine of not having the physics and collisions independant of fps. That is very evident when comparing collisions on 30 fps to collisions on 5 fps for example, low fps causes problems such as the pinball mode while the exact same collision on high fps would always results in a quick halt with no air time.

If thats the case its more likely got to do with the processing speed. There's a possibility the processor is dropping "frames" dealing with the physics speed. Collision detection is one of the most difficult things to master with making your own engine, and people need to remember that most of the engines for the games such as Quake, Half-Life, and UT (and all of their derivatives) have people who've been programming engine physics for 20 years.

They've fixed it though, which is great news. Now I don't have to be scared about hitting a rock at a bad angle in a tank.

[R0adki11 3949]

Quote[/b] ]Havok doesn't instantly mean 'ragdoll'. Havok physics are used on vehicles, objects and flight models as well.

Physics aren't everything in a game, there is game play, mp, sp, all add to how good a game is, you cant have everything in agame otherwise we wouldn't have such hug lanscapes otheriwse u would need faster machines etc

[Vassago 0]

Roadkill, I've been in game development for 8 years, man. In a game like this, physics are a HUGE part. And they directly affect gameplay. Flying, driving, collision & ballistics are ALL handled with physics. For a game that touts so much realism, it's core physics simulation is completely fubar. So honestly please, don't try telling me 'you can't have everything'. You do what needs to be done for the type of game you're making.

How about a boat simulator that doesn't have proper water simulation? It wouldn't feel right, would it? Neither do the physics in ArmA. It's improved over OFP for sure, but they're not solid.

[Metal Heart 0]

If thats the case its more likely got to do with the processing speed. There's a possibility the processor is dropping "frames" dealing with the physics speed.

Yes, that is exactly what I meant, that the collision/physics don't run on a constant Hz but skip cycles, or more correctly, have varying intervals on the cycles depending on load other than physics.

[Big Dawg KS 6]

Quote[/b] ]i was only pointing out that bis has obiously put a incorrect value when vehicle hits rock, Instead of 5km jump  value, there should be 5m

...

So basic math for BIS

The only obvious thing is that you have exactly zero understanding of basic physics programming If it were only a matter of a single incorrect value the problem would be consistent.

I think there is a fundamental design flaw in the engine of not having the physics and collisions independant of fps. That is very evident when comparing collisions on 30 fps to collisions on 5 fps for example, low fps causes problems such as the pinball mode while the exact same collision on high fps would always results in a quick halt with no air time.

Yes that is related to the problem, but for some reason, even with a decent framerate (not that I usually have a good framerates), the collision detection is not very precise. Objects with small geometry, such as a grenade, will sometimes entirely submerge into the ground or pass through surfaces before the collision detection kicks in and pops it back out (at which point the speed at which it is bounced back is ridiculously high).

[xnodunitx 0]

Well then it can be solved rather easily, since it does indeed seem that smaller geometry has issues and not big ones, that would explain perfectly why the treads sometimes sink into the ground or other objects, causing it to get stuck and then be shot up.

Provided it is this simple and its probably not..the geometry just needs to be larger, however I would think the tank geometry has a single box for its treads, not an actual tread section..hmm..this will be interesting.

But it could be one of the answers, its worth a try atleast, if an OFP addon could be ported into the game just to test this theory then we might be able to get our answer.