Comments on: Hard Object Weirdness (or Don’t Count On Bugs)

By: Benjamin Supnik

Benjamin Supnik — Thu, 26 Feb 2009 08:43:00 +0000

I think the seaplane behavior is a different limitation – we don’t track the ground and objects at the same time.

By: Photos

Photos — Thu, 26 Feb 2009 03:56:00 +0000

is it the same problem while beaching a seaplane or try to use a slip ? Any ways for an improvement to avoid huge collision behavior ?

Olivier

By: Benjamin Supnik

Benjamin Supnik — Wed, 25 Feb 2009 15:11:00 +0000

Anon: neither should have changed in 930.

Re: flying under, the author must use ATTR_hard_deck to allow this. ATTR_hard_deck worked in 922 and works in 930, so if you cannot fly under the arch, the author has not yet taken advantage of this (relatively recent) capability in the sim.

Re: the helipad, this is a design limitation, not a bug. There is no way to specify a helipad on top of a building right now. 930 does not add this capability.

By: Anonymous

Anonymous — Wed, 25 Feb 2009 14:46:00 +0000

There is Grand Arche, part of Bretrand's Paris scenery. I tis a hollow building and it has a helipad at top. It had two issues: (i) Place a heli to that helipad, and it ended in the basement. (ii)one cannot fly thru the opening in the building.

Has any of these got better in 9.3?

By: Benjamin Supnik

Benjamin Supnik — Wed, 25 Feb 2009 09:26:00 +0000

Alejandro, the issue is this: since X-Plane's collision logic is only vertical, if you very gently bump into the "top" of a door (modeled by the horizontal top of the door being a hard polygon) the collision strength is measured by how far below the door you are vertically. This results in a huge collision even if your plane is moving at one mile an hour.

In an "any-direction" collision model, we would test collisions in _all_ directions...allowing for a very small collision of the nose of the aircraft with the surface directly in front of it, instead of a huge one with the surface above it. That surface in front of it would of course have to be vertical.

In other words, the issue of vertical surfaces is closely tied to the issue of any-direction collision detection; without this feature, we correctly determinet that there IS a collision but INCORRECTLY judge how violent that collision is!

By: Alejandro Garcia

Alejandro Garcia — Wed, 25 Feb 2009 06:30:00 +0000

Ah...I whish x-plane were GTA4...
No, seriously I don't really understand why would we need a vertical hard surface. I'm sure there are lots of implementations but right now I can't see some.

If we talk about collisions, you can add a little horizontal surface so the plane will collide and it'll look as if the vertical object was the problem.