[Killer Whale]

Maybe they don't have artificial gravity, maybe they're just that heavy
(j/k)

[IronBeer]

Anywhere the acceleration due to gravity is nonzero

Which is practically nowhere statistically speaking

So there you go, following on from that, there is pratically nowhere in space where you don't have weight.

There is no point in space where an object will have zero weight, as gravity can act at an infinite distance. The force of gravitational attraction may be hilariously minuscule, but it will never be zero. But I'm sure you already knew that.

On-topic, I actually wouldn't mind having reasonable values for the mass/weight of the various ships, since I'd want to see/devise some actual specifications, and mass would be a big one.

[Unknown Target]

Aren't the 3D models simplified into boxes for collision detection?

To answer your question, because I didn't see anyone else do it; in Freespace 2, no, they're not, it's all calculated off the vertex normals. That's why it's so important to weld all your verts; too many and the game freaks out.

[Alan Bolte]

I was wondering how that worked. Is that... efficient at all? My understanding is that most games' collision detection requires the artist to create separate collision models which must be composed of convex hulls.

[Thaeris]

How does a vertex have a normal? A vertex is merely a point in space which defines one of two points which make a model's edge. Rather, do you mean that FS uses the normals from a surface/face of the model's geometry to calculate hit impacts?

OR do you mean to say that FS isolates adjacent groups of vertices and computes a normal from those groups... Yes, that must be it.

[The E]

To bring an end to this confusion:

Yes, FSO does not use simplified collision models.
No, that is not in and of itself a performance risk, because we are using something called BSP trees to subdivide the geometry. This means that, when a collision check is performed, the engine uses a succession of abstractions starting from the models' bounding box, to a subobject's bounding box within that, to a specific polygon's bounding box until it gets to the specific polygon.

Now, this tactic has worked perfectly well for retail-era models, it's only recently that it has shown its age, given the level of detail present in modern models. It is possible to do collision models (seriously, the framework is already there in the models, and in particular the $nocollide_this_only submodel property), but it's not something that is in wide use, or really tested that much.

[redsniper]

How does a vertex have a normal? A vertex is merely a point in space which defines one of two points which make a model's edge. Rather, do you mean that FS uses the normals from a surface/face of the model's geometry to calculate hit impacts?

OR do you mean to say that FS isolates adjacent groups of vertices and computes a normal from those groups... Yes, that must be it.

No, you probably won't hear about them in geometry class, but they do exist in the 3D graphics world, have been around for a while, and are quite useful.

http://en.wikipedia.org/wiki/Vertex_normal

[karajorma]

There is no point in space where an object will have zero weight, as gravity can act at an infinite distance. The force of gravitational attraction may be hilariously minuscule, but it will never be zero. But I'm sure you already knew that.

Well there may be a point where they all balance out to 0.

[IronBeer]

There is no point in space where an object will have zero weight, as gravity can act at an infinite distance. The force of gravitational attraction may be hilariously minuscule, but it will never be zero. But I'm sure you already knew that.

Well there may be a point where they all balance out to 0.

Ok, yeah. Didn't consider that. Such a force balance would be the most tenuous of tenuous systems, though. Any shifting of any of the objects involved in that system would result in nonzero sigma-F.

I don't mean to be argumentative; I'm pretty certain we both know what we're talking about, I just want to state the technicalities. Shall we let this drop?

[Cyborg17]

Actually, can't it be zero?....  If plank length didn't exist I would agree with you, but if plank length exists, and gravitation is a distortion of space time, there should be a point where the change caused by the distortion is less than plank length.   Wouldn't that cause the distortion to cease?

EDIT: I'm just intrigued by the possibility you brought up.  I don't actually know what anybody is arguing for.

[BengalTiger]

So plank length is the smallest possible distance the computer running our universe is able to calculate with it's current precision?

[LordMelvin]

So plank length is the smallest possible distance the computer running our universe is able to calculate with it's current precision?

No, plank length is 14 feet, at least for the one-by-twelves we've got out in the garage. Planck length, on the other hand, is what you said.