Non-Static Collsions
Blitz3D Forums/Blitz3D Programming/Non-Static Collsions
| ||
| Ive read a fair bit about instability/unreliability of moving entity/moving entity collisions, and was wondering how unreliable they are, what criteria are more/less reliable, why this happens and any possible workarounds. My game involves a lot of collisions with meshes, sprites, and pivots. All of which may be moving, rotating and/or static at the moment of impact(s). |
| ||
| Well only ellipse to ellipse works when both entities are moving and that comes with caveats about how many collisions are reported and by which entity. Ellipse to box did not work properly last time I tried, which I think was 1.87, though sphere to box is ok. |
| ||
| Well I have been working on a space combat game for a while and it uses sphere to polygon collision between weapon particles and ships. Even though both can move quite quickly, I have not noticed any major problems with collision detection. I think perhaps frame rate could be important, because it helps determine the amount that the objects move by before UpdateWorld is called. Maybe high frame rates could reduce the inaccuracies? |
| ||
| Well Ive not "noticed" any problems yet, I just wanted to be sure I knew what I was lookin for. Thanks. |
| ||
| Problably you'll be able to fix missing collision behaviour (eg. when 2 meshes are moving towards eachother!) when you add an "updateworld(0)" after every parties moves. I mean, when you have 2 spaceships and several asteroids and you only want to check if a ship collides with an asteroid, but don't care if asteroids collide correctly with other asteroids, or ships collide correctly with other ships, you would have to move all asteroids, then updateworld(0) then move all ships and then do the updateworld() as usual. |
   |