TestTubeGames Forums

Speaking of changing the range, the range input might be extended (a longer bar may popup above the planet/star button) for greater precision.

Or maybe we could have sliders that change the lower and upper limits of the slider, and we could have sliders to change the upper and lower limits of those sliders, too, and so on, until we have enough sliders to start a diner!

Or we just use a textbox input

exfret wrote:Why is it such a common happening that when someone replies, someone else comes in and finally replies to a completely different comment posted months before

Must be that distracting picture of Ian Malcolm above

Indeed, welcome Gravity Coil! I agree with you, I'm always looking to break down restrictions when at all possible. And right now, you can't easily add a planet (or 'moon') with a mass less than 10. The slider doesn't support it. Nor can you easily pick precise values. So maybe a text-input would be a way to go. Or a slider, combined with a text-input (for people who are so inclined).

But I'm with you all. There should be a way to pick extremely tiny masses, extremely big masses, and everything in between.

A Random Player wrote:Speaking of changing the range, the range input might be extended (a longer bar may popup above the planet/star button) for greater precision.

You should have a button that opens a popup box where the gravitational equation can be changed. Like this :

Original equation: F(M,m,r)=G*M*m(r^n) where n is the gravity exponent (This isn't complex enough for certain conditions, such as General Relativity, where the equation is more like G*M*m*(r^-2+n*r^-4), where n is a very small constant, usually related to the speed of light.)

Custom equation example: F(M,m,r)=G*M*m*j(r) where the j function is a custom function, like j(r)=r^n+a*f(b*r+c*theta)+1 (f is a common math function, such as sin, cos, tan, log, ln, etc.; a, b, and c are variables. "theta" is the current angle between the smaller body and the larger one, with respect to the X axis.), or j(r)=(r^a+r^b+r^c+...)/n, where "a, b, c, ..." are different exponents and n is the number of terms in the equation.

I've been looking all over the Internet for something like that, but to no avail (seems that no one made custom gravity simulators). Also, you need to add an option to show/hide the gravity field (maybe as a color-coded gravity map) when the simulation is paused. That will make it easier to put objects in orbit, especially for r^100 and r^-100, where the gravity well is extremely steep inside/outside a certain radius. Could you please do this?

Stargate38 wrote:You should have a button that opens a popup box where the gravitational equation can be changed. Like this :

Original equation: F(M,m,r)=G*M*m(r^n) where n is the gravity exponent (This isn't complex enough for certain conditions, such as General Relativity, where the equation is more like G*M*m*(r^-2+n*r^-4), where n is a very small constant, usually related to the speed of light.)

Custom equation example: F(M,m,r)=G*M*m*j(r) where the j function is a custom function, like j(r)=r^n+a*f(b*r+c*theta)+1 (f is a common math function, such as sin, cos, tan, log, ln, etc.; a, b, and c are variables. "theta" is the current angle between the smaller body and the larger one, with respect to the X axis.), or j(r)=(r^a+r^b+r^c+...)/n, where "a, b, c, ..." are different exponents and n is the number of terms in the equation.

I've been looking all over the Internet for something like that, but to no avail (seems that no one made custom gravity simulators). Also, you need to add an option to show/hide the gravity field (maybe as a color-coded gravity map) when the simulation is paused. That will make it easier to put objects in orbit, especially for r^100 and r^-100, where the gravity well is extremely steep inside/outside a certain radius. Could you please do this?

I am okay with this.

I don't care how weird to use it is, I demand you add such features.
F =G*M*m*f(r)
Then you are allowed to change f(r). By default it's r^2, and as of right now you can change it to r to the power of any real number.
However, this would allow you to do so much stuff. Like, r^r. Or 2r. or even r+r^2
The possibilities are endless!

I wonder what orbits would look like with an r^(1/r) gravity law. At large distances, this would approximate r^0 (you know what that looks like), but at small distances, such as r<50, it would act different. With 0.5^r, you probably couldn't make a stable, non-circular orbit, like r^-3.

Andy, stop all progress on the electric shocktopus. This is a neccessity!

I wonder how sin(r) would do.

How would the integrator work? Are you sure this is even possible to do this accurately? Wait, how does the integrator in the current Gravity Simulator even work anyways? I've always wondered that...

robly18 wrote:Andy, stop all progress on the electric shocktopus. This is a neccessity!

I wonder how sin(r) would do.

Interesting. So I just make it quasi-wolfram-alpha style -- where you can plug in any sensible math string? That's quite some... 'power'... you'd have. I like it. Hopefully it won't be too much of a pain to implement.

As for the integrator, exfret... it works through Runge Kutta 4 methods. In laymen's terms, it's just a very good way of figuring out where the next spot of an object will be, if you know the forces on it. So these extreme force laws shouldn't present any problem. As long as we can calculate the forces everywhere (no dividing by zero problems, say... hrm, add that to the list of challenges with no-holds-barred force laws), the integrator would work.

Now, whether the orbits would be stable... or if they'd be crazy, that's another question entirely. (I'm going to go with crazy)