Suggestions! (new)

[exfret]

I LOVE this! It adds in a way to make a force other than toward and away, which I've always wanted, and it makes perfect sense, too. It's just like the imaginary plane: if -1 is away and 1 is towards, then i and -i would be in between!

[robly18]

I LOVE this! It adds in a way to make a force other than toward and away, which I've always wanted, and it makes perfect sense, too. It's just like the imaginary plane: if -1 is away and 1 is towards, then i and -i would be in between!

Well, it makes sense, until you think about it better.
I mean, it works on a 2d plane where there are 2 axis just like in the complex plane, but if we were on a three dimensional space, this would no longer apply.
But I guess it works as a way to make transversal gravity, though I think physically a separate box for transversal force would make more sense, but with this you could do it simply by adding i * transversal force.

So I don't know. I'll let Andy be the judge, and I think I know who he's gonna rule for.

I don't oppose per se, anyway.

[exfret]

Think about it 'better'? You've figured out a way to measure aspects of one thing over another that shows which one results in the outcome that fits every person's values more? Tell me how! (You mean, think about it 'more', right?)

Also, why doesn't it make sense in 3D? Most of the time, i and -i come together anyways (in fact, they're exactly the same except for the fact that there are 2 separate 'instances' of them), so that would be two forces (left and right) when they both pop up at the same time anyways, meaning it could just be the same way in 3D (force goes in all directions perpendicular to the 'regular' force line) which leads me to a question: how will you deal with multiple force values? I mean, you won't just take ^0.5 as just the principle root, will you? With the multiple i directions, you could just make it so that any expression that results in a complex number and it's conjugate could just result in having only the real part taken for the force, because the imaginary forces would cancel that out. That makes sense... You could also be able to specify specifically only i or -i and not it's conjugate, and the fact that all the i terms would be together would allow for separateness in the force box so you can differentiate between real and imaginary forces. (While the GSim integrates!)

Also, here are some more variables that you could let us use:
Time
Random stats (eg # planets)
Angle
Allow us to modify the m1m2 part and give us the ability to make acceleration laws so it's easier to imagine
Allow use of piece wise functions
Tell the GSim to use different force laws for different objects...

I'm going to stop there before I go into overdrive mode...

[robly18]

Also, why doesn't it make sense in 3D?

Because in which direction would it go? Parallel to the object attracting you? Let me know how that goes: you've an infinity of directions to choose!

Most of the time, i and -i come together anyways (in fact, they're exactly the same except for the fact that there are 2 separate 'instances' of them), so that would be two forces (left and right) when they both pop up at the same time anyways, meaning it could just be the same way in 3D (force goes in all directions perpendicular to the 'regular' force line)

Then complex valued forces would be, quite literally, null.

which leads me to a question: how will you deal with multiple force values? I mean, you won't just take ^0.5 as just the principle root, will you? With the multiple i directions, you could just make it so that any expression that results in a complex number and it's conjugate could just result in having only the real part taken for the force, because the imaginary forces would cancel that out. That makes sense...

Pretty much.

You could also be able to specify specifically only i or -i and not it's conjugate, and the fact that all the i terms would be together would allow for separateness in the force box so you can differentiate between real and imaginary forces. (While the GSim integrates!)

Also, here are some more variables that you could let us use:
Time
Random stats (eg # planets)
Angle
Allow us to modify the m1m2 part and give us the ability to make acceleration laws so it's easier to imagine
Allow use of piece wise functions
Tell the GSim to use different force laws for different objects...

I'm going to stop there before I go into overdrive mode...

I'm okay with time, mostly because then it would be possible to make a cos(t) function to allow for oscillating forces.
As for random stats, meh. Angle, would kinda break the relativity principle but hey, this is a video game!
Allow us to modify the m1m2 parts? Uh, I think it's kinda fine at the moment.
Piece wise functions? Explain?
As for different force laws for different objects... This is just shifting away from real life physics by this point.

[exfret]

3D would be pretty much the same result of 2D: nullification/cancellage out. Also, piecewise functions would be like (r<2)(t>1)*r^2, where an expression like x>n returns 1 if true and 0 if false. And, since when have we not diverged from real-life physics?

[robly18]

3D would be pretty much the same result of 2D: nullification/cancellage out. Also, piecewise functions would be like (r<2)(t>1)*r^2, where an expression like x>n returns 1 if true and 0 if false. And, since when have we not diverged from real-life physics?

Well, whacky f(r) laws are not necessarily a part of our universe, but are certainly possible. However, complex values for forces suddenly turning into a transversal force is very arbitrary.

But this is just my close-minded rants. Add it if you will, but I formally would prefer it if complex values would cancel out.

Keep in mind though, this is just my opinion. No friends are being lost if this happens! It's Andy's game, he does whatever he wants with it.

[wtg62]

Hmmph... I think a Rand() function (That gets a random number between 0 & 1) would be good.

Why between 0 & 1?
JavaScript does it this way, and it's actually quite easy to allow the user to select a range. All you gotta do to get your maximum is multiply the number, and to get your minimum, add to it.

i.e. (Rand()*9)+1 will give us a random number between 1 & 10.

Then we could do r^-((Rand()*9)+1) and voila, jerky motions and chaos! A force law between r^-1 and r^-10 (Be warned though, it won't select JUST whole numbers).

[exfret]

Wait, so would it choose a new random value every update? What about having the force value randomly go up and down. It could vary. Oh, I guess just adding in a time variable would allow that. Anyways, how could it possibly run if every update you have a new force law to deal with?

[testtubegames]

Interesting thoughts here.

Robly, I'm completely on-board with the fact this wouldn't work in 3D, since you wouldn't have a defined direction. And it is a bit strange when it comes to physics-representation. But I'm curious to give it a try. At the very least, it is just a way to add in transverse forces. Down the road, maybe the two forces would become decoupled... so you define them each, or with a vector, instead of using complex numbers.

exfret, as to your point about 'which root would you use', the answer is the principal root. Right now, if you've got a force law of r^.5, and you are 16 units away, the force is 4. And if you were instead under the force law (-r)^.5, the force would be 4i. That's just the simple way of doing it, and not without physical justification (you often choose one of the roots, or give a physical explanation of why the other is ridiculous: "you won't *repel* in an attracting force!")

Time makes sense to me as a variable, though it'll be harder/more computationally intensive to implement. Doable, though, at least for <10 planets. (Of course, things would go wacky quickly if you used t linearly... ) Random is possible, too, though I'm not sure it would be too useful. As exfret was asking, would it recompute every calculation? Seems we might lose the 'universal' law aspect of the sim.

As for the other variables, I'll keep them in mind. Angle could be interesting (though of course, completely physically irrelevant... since what is an angle in the real world?) I imagine I'll hold on to these, and add some in later if/when I get a chance/have people storming the gates for these features.

[robly18]

I just realized something.

If we had the angle, we would have the position relative to an attracting object using a few trigonometric functions.

I'd say it's safe to assume that, if this was added, the whole notion of relativity would be going out the window for the simulator.