Gravity behaviours unnaturally with any setting....

Hunnenkoenig

I looked on the forum and FAQ, but found nothing.
There is probably a solution, but I couldn't find it.

I wanted to experiment for my next game idea and made an actor and placed it in 3 instances on the screen like \___/

I dropped a ball and it rolled down on the left wall, then up on the right wall and back again. So far is everything OK

Now the problem is, the ball keeps rolling even if it should already be standing. It rolls on the horizontal line back and forth very very slow, but it seems never to stop or just after an unnaturally long time.

I tried every possible setting with density and friction with values from 1 to 10 000 on the wall and on the ball as well, but no difference.

I tried to play with the value of the gravity, but it only affected, how slow or quick the ball rolls.

Any thoughts on this? Do I have to add something to the scene?

Hunnenkoenig

Nobody can help with this?

jonmulcahy

if you increase the friction on the floor, and decrease density on the actor, it should roll to a stop i think. I don't have GS in front of me to test it though.

try a friction of 100 on the floor, and density of 1 on the ball? i think the higher the friction, the quicker lower density objects will loose their momentum

Hunnenkoenig

It's a bit better with your numbers, but it still rolls and rolls and rolls very slowly until it can't be roll upwards anymore and it stops at the left wall.

I tried with gravity 10, 100, 200 always the same.

rebump

Build a test tool app where you can alter the values (like a slider) while the test app is running. It takes a little longer but it sure beats edit/run/edit/run/edit/run...

I do this all the time when testing out firing rates/times/distances, relative movement speeds of actors for chase, rotation tests, etc.

Then you have tools you can always re-use later for the next project.

Hunnenkoenig

good idea :-)
I will try to get the best results.

None the less, in this state gravity is crap in the physics engine.

I don't know, if the devs can alter it or if this is a 3rd party thing, but I hope, it will be updated somehow.

rebump

I think you just need patience. I've seen some good results in some of the games. I know FMG has some smooth moves and I know he takes his time iteration after iteration.

Thusly my suggestion of a test tool app to speed the process.

The Box2D engine that GS is using is supposed to be highly rated. GS doesn't use it all but it uses enough that there is a lot to play with to get it down how you want it.

Hunnenkoenig

I don't know if 3 switches/attributes are enough to have good physics...

I tried every possible settings and it's always the same. There I can't see any possibilities to make it realistic. But I will try.

And I don't want to sound harsh, I jsut say what I think, reacting to your text, but if you make a software, which is supposed to be used 99% for physics games, then it is not enough to "not use all, but use enough"... then I expect, it uses everything, if that is all, what I can create with it.

If you understand, what I mean :-)

rebump

I do understand but alot of the features GS is not using is higher end stuff like skeletal/bone segment linking movement stuff and some other stuff that is more 3D oriented. I'm sure more will trickle in over time but given that GS is 2D, I think the bulk of what is needed is there. Sounds like there is more collision stuff that could be added and hopefully that will makes its way in eventually.

Hunnenkoenig

I see.

I don't know what is possible and of course, we don't need 3d stuff for now (later it would be nice), but what we can use, I want to use :-).