So, Pausert & I were having this argu^H^H^H^H discussion about the physics in the game, and we made a bet on how it works. So, question for devs on how the game's physics works: Are there different ship classes (light, medium, heavy), that all have seperate physics models, or do all ships conform to the same physics model?

Also, do the different ship models (Cent, Valk, Maud) have different mass distribution models, that affect their rotation?

who was on what side of the bet, so i can laugh my head off at the one who thought there were different ship classes?

Our physics model is consistent. There are no special cases made for ships, just different parameters, such as top speed.

Cool. What about the mass distribution for angular acceleration, is it considered to be an even spherical mass distribution for all ships, or do the different ship models have different mass distributions?

to test that is simple, turbo really fast, turn your camera, be in physics mode, and see if your ship slows down with the nose pointd somewhere other than right where you are looking

Actually, my feeling was that the Atlas Type X handles suspiciously well when compared to, say, an scp with a similar weight (or lower) weight. And in addition, it seems like the Maud turns much better than other ships with similar stats when weighed down. I figured that was the advantage of the maud - it was a light ship that had a large cargo bay. Instead, now I'm a bit confused as to the purpose of the maud. Who cares if a maud is light? Its weight is always at least doubled when loaded with cargo. So what is the point of the maud now? Can anyone see ANY use for this ship?

EDIT: Beolach, I apparently owe you some credits :)

Irrelevant now since all the updates, but this has been tested before ;)

http://www.vendetta-online.com/x/msgboard/1/2126

Someone feeling up for a challenge wanna try this on all the new ships heh?

Well, there are game control settings that will affect that, as well as the game's physics engine. Try "/set controlresponse 0.1", and watch as it takes forever for your ship to turn. But as long as you keep the controlresponse constant throughout all your tests, they would still be a good indicator of how fast the ship turns.

And to answer my question about the mass distribution, the tests by themselves would be somewhat lacking. To answer my question, it would require doing a lot of math for several different ship models & stats, with the assumption of an even spherical mass distribution, and then testing all of those ships to see how well all the tests match up with the calculated results. If they all match up well, then it means all the ships use an even spherical mass distribution, and if they don't match up, then it means they have their own mass distributions.

I'm too lazy to do that. It's easier just to ask the Devs. ;-)

Well, if you look at how the ships are ranked, some are ranked higher in linear acceleration than they are in rotational acceleration. This would not be the case if all ships had spherical mass.

I will bet (with real credits ;)) that the mass distribution is also consistent. And that it is homogenous and based on the geometry of the ships themselves.

(This is another way of trying to pick the devs' heads.)

@Beolach : There is very little math involve in computing a ship's moment of inertia, it is a simple integral that can be done once and for all with a computer if you want super duper accurate numbers. If you want to do it by hand, you can always approximate (use a rectangular parallepiped : they are not far off at all).

@Phaserlight: I don't think it's that simple. They could rank differently in linear acceleration than in rotational acceleration, even if they all had spherical masses, if they had different diameters and/or different spin torques.

@softy2: How do you mean "mass distribution is also consistent"? I would interpret that to mean all ships had the same mass distribution, but then you say "and based on the geometry of the ships themselves", which would mean different ships would have different mass distributions, which would mean to me that they were not consistent. Or do you mean that the density would be consistent throughout the shape of the ship?

As for "There is very little math involve in computing a ship's moment of inertia" - if you want to try it, go ahead. I'm still way too lazy. Also, we don't have enough information to be sure we calculated it accurately. We'd need to have the shape that the mass is distributed through (possibly a sphere, or the ship's graphic model, or a near approximation of the ship model, or something else), and we'd also need to know whether or not the density is consistent throughut the shape, or if some parts of the shape are more dense than other parts.

Anyway, my bet is that the moment of inertia is calculated off a near-approximation of the ship's graphical model, and that the density is mostly consistent throughout the shape, but that some volume of the shape is designated as the cargo hold, and the density of this volume may be different from the density of the rest of the shape, depending on how much cargo mass in in the cargo hold. There may also be some volume of the shape set aside for the weapon ports, and the density for of those volumes would be based on the mass of any weapons equipped. But that's just my guess.

@Beolach; true, I hadn't considered it could be a simple function of diameter. Spin torque was the same on all ships, however, since this test was done back in the day where different engines with unique stats were equippable. I just equipped the same engine on all the ships.

@Beolach :

"Consistent" in the sense that the devs did not cheat and put in some arbitrary values. Sorry if my post is a bit too mathy. Let me reparse it : the devs use a *Homogenous* (constant) *mass distribution* (i.e. density as a function of position) and compute the the Moment of Inertia, integrating over the geometry (the shape) of the ships.

This means that there is no "empty spaces" inside the ship (e.g. cargo hold), since \rho(x) = constant \forall x=inside ship and \rho(x)=0 otherwise.

[EDIT : The density is computed from the total mass (with loadout) / volume of the ships. So there is nothing hidden here too.

What is hidden, is of course, the damping factor of each ship. That there has to be a damping factor is obvious : else you will just spin your ship to ungodly rotational speed if you simply apply torque constantly. This factor can be tweaked by hand by the devs, since there is nothing in the ships stats which you can derive this value from. In fact, this is a cheat :), since in space there is really no damping from any atmospheric friction. This factor will govern how well a ship "snap" into position after a turn.]

I have computed the M.I. for some of the ships too. Just use a formula for the rectangular parallepiped and plug plug plug.

And *that* is what I'm talking about. That is how a medium ship that weighs more and has less thrust than a prom can *feel* lighter and more manueverable. I just couldn't explain it well. This is perhaps due to the fact that I am an utter, utter, imbicile and, oh wait, gotta go, time to change my drool cup.

@softy2: Heh, it wasn't the "mathy" parts that confused me. Mostly it was my not having any sleep for too long, but what I was having trouble with was figuring out what the adjective consistent applied to, an individual ship, or all the ships together.

Can you prove that the moments of inertia you calculated are correct, though? How do you know that ρ(x) = constant for all x inside the ship?

@Beolach

Well I can't prove they are right, which is why I am hoping we can get some enligtenment from the devs themselves :).

so if i follow, asymmetrical ships would be nightmarish to pilot...

Tell that to Han Solo.

I hope not cuz it would be so cool to see this http://img.photobucket.com/albums/v178/Dimpfinator/Anubis.jpg

or pilot one of these
http://www.myklittle.co.uk/stuff/elite/Itani_Hammer.jpg
in game.

BTW: controlresponse ONLY affects mouselook. :P