3D physics (motion) libraries

00000000000000000000000000000000
0
Anonymous Oct 31, 2004 at 20:26

Hi, just like the subject suggests.

I’m having difficulty in tracking down a suitable library that will, preferably in c++, given a location, velocity, azimuth and angular velocity (in three dimensions), and given a mass and inertia, then given also a time interval plus an acceleration and angular acceleration, a new “state” will be produced.

found several references to books, but no actual code…

anyone got any references? preferably to open source or c++ code…

l.

8 Replies

Please log in or register to post a reply.

7543b5c50738e23b200e69fe697ea85a
0
NomadRock 101 Nov 01, 2004 at 01:26

You can do that easily with a few lines of code. That is extremely basic Newtonian physics. Read an intro to physics page, or preferably your school book.

2b97deded6213469bcd87b65cce5d014
0
Mihail121 102 Nov 01, 2004 at 05:16

check out the collection at http://www.gamels.net

00000000000000000000000000000000
0
Anonymous Nov 01, 2004 at 08:11

hi,

well i am sure that it _is_ a simple task - for someone with the knowledge already burned into their brain by repeated exposure to it. just a piece of advice: in future, when you come across someone who asks such a question again, try to answer the _question they ask_ without clearly demonstrating the difference between your knowledge and theirs, and then having accentuated that gap, then using it as a justification to not actually answering their question.

l.

00000000000000000000000000000000
0
Anonymous Nov 01, 2004 at 08:12

@Mihail121

check out the collection at http://www.gamels.net [snapback]13508[/snapback]

BRILLIANT, thank you: a couple of clicks on the physics section of that site led through to www.ode.org.

thanks v. much, you’ve been very helpful.

l.

00000000000000000000000000000000
0
Anonymous Nov 01, 2004 at 08:17

@Luke Kenneth Casson Leighton

@Mihail121

check out the collection at http://www.gamels.net [snapback]13508[/snapback]

BRILLIANT, thank you: a couple of clicks on the physics section of that site led through to www.ode.org.

thanks v. much, you’ve been very helpful.

l.

[snapback]13512[/snapback]

http://www.ode.org/ode-0.5-userguide.html

this is so cool. it covers ball joints, sockets, pistons, hinges, friction, torque, the lot.

[and it doesn’t heavily mention physics books in lieu of actual useful information,
and it doesn’t mention going back to school. i do hope you won’t do that again,
mr jesse coyle]

7543b5c50738e23b200e69fe697ea85a
0
NomadRock 101 Nov 01, 2004 at 10:49

Sorry, I was a bit harsh. However you must realize that with the level of the question you origionally posted, I was not led to believe you wanted much more than s=s0 +v0t + a0t\^2 and the like.

I do apoligize, I judged you incorrectly, but I will not promise not to do it again. I am here not merely to teach others algorithms, but to teach others how to find those algorithms themselves. If someone who truely only needs basic physics cant be bothered to use google, then they deserve a cold shoulder.

That said, I am glad you found what you were looking for.

00000000000000000000000000000000
0
Anonymous Nov 01, 2004 at 22:03

@NomadRock

Sorry, I was a bit harsh. However you must realize that with the level of the question you origionally posted, I was not led to believe you wanted much more than s=s0 +v0t + a0t\^2 and the like.

I do apoligize, I judged you incorrectly, but I will not promise not to do it again. I am here not merely to teach others algorithms, but to teach others how to find those algorithms themselves. If someone who truely only needs basic physics cant be bothered to use google, then they deserve a cold shoulder.

That said, I am glad you found what you were looking for.

[snapback]13516[/snapback]

dear jesse,

on virtually every new forum i post, i get into a bun fight with the first people i meet: invariably over misunderstandings. the fault is mine - for not doing enough research into the forum before posting.

google led me to this site (amongst others) - in 30 mins of searching i had 15 bookmark references (one of them being devmaster.net) and then once i was kindly referred to that, i found another 5.

ode is exactly what i was looking for: it is incredible. not only does it have the libraries i believed would be around if i searched long enough but it actually turns out to hook into OpenGL on your behalf!

what i mean by that is that you define basic objects (cylinder, sphere, rectangle) and the relationships between them (joint, bend, piston, collision) where “ground” is just another of the objects, then you call the “sim loop” function and voila, up pops a 3D real-time view!

there’s a basic example of a “car” - test_buggy.cpp - which has 3 narrow cylinder wheels and a rectangle block chassis (made of lead as far as i can tell from the behaviour of the “car”!).

it just… _so_ far exceeds my wildest expectations.

l.

7543b5c50738e23b200e69fe697ea85a
0
NomadRock 101 Nov 02, 2004 at 04:18

Yeah, I am familiar to ODE. It is quite good. Especially since it was made more open recently.

Well, at any rate, confusion aside, I am happy with your success. I hope we will be seeing a fun demo in the future showing off the dynamics a physics engine provides.