The Nineteenth Byte

no, its pretty cool

@AshwinGupta i like it

Geobits

You wouldn't think some simple side-scrolling lines would be so damned disorienting, but wow.

As for the background, I am keeping the light rays, but I'm going to re-space them obviously.

5:02 AM

@Mego now that I have the magnitude of the velocity, but how would that help me if I don't know the direction?

It's the first thing I've programmed in JS/Html so yeah still learning

@Maltysen lol serious? I honestly thought it is kinda ugly.

@AshwinGupta yeah the trail effect is pretty nice

Ok ty

I'll keep it in their then

@Downgoat you know the direction

its tangent to the orbit

Tomorrow, I'm fixing the background, then it will be saturday so I'll have plenty of time

5:03 AM

oh

ugh, maths

which means its normal to your position vector

to hopefully program the missile hit detection, then some of the alien ai

next week I'll do menus/weapon selections. Finish the ai up

kepler's laws!

and finally I'll do in game power ups and points/health

@Maltysen Math is the worst ;_;

5:04 AM

@Downgoat kepler's laws and newton's derivations are really cool

speaking of Kepler, did Nasa ever find out the cause of the Kepler (Satellite) failure that caused it to go emergency mode?

Anonymous

@Downgoat Equation of an ellipse is x^2/a+ y^2/b = 0. Implicit differentiation, equation of tangent line is x*x_0/a + y*x_0/b = 0. Get your line, find the angle.

@Mego wait can't he just say that the velocity is normal to position vector?

@AshwinGupta halp what is kepler satellite?

Anonymous

@Maltysen Yeah he could do that too

@Maltysen =/ its for finding planets... uses transit method to locate exo planets

Kepler

5:06 AM

@Downgoat do you know how to find slopes of lines and then slopes of perpendicular lines?

@Maltysen yeah, except I don't want to do this in a very inefficient / slow way

opposite reciprocal then * -1 isnt it?

For perpendicular ^

@Downgoat no, that's one of the fastest ways, except for going fully closed form

@AshwinGupta yup

Is their a better way to do it that I've missed or not learned?

Anonymous

You could just used the closed-form

5:08 AM

@Downgoat find the slope of your position vector, and then find the slope perpendicular to that, and that's the slope of your velocity vector

Anonymous

You have your ellipse characterized by a center (X,Y) and a pair of semi-axes (a,b) (a is semi-major, b is semi-minor)

@Mego he's doing restricted 3 body though, which as you said above, is very hard

Anonymous

Your position on that ellipse is characterized by the equation x^2/a + y^2/b = 0

cuz of moon

Anonymous

@Maltysen The closed-form solution works for discrete time, as long as you re-calculate the potentials every iteration

5:10 AM

oh i see what you mean.

Anonymous

Wait this is dumb

Anonymous

This can be done super easy

Anonymous

Place your sun at the origin of your coordinate system (0,0,0)

^^^

@AlexA. ...I have so many conflicting feelings about this. :P

Alex A.

5:12 AM

hahaha

I can imagine

Anonymous

Your planet and moon start with initial position vectors xp_0 and xm_0, and initial velocity vectors vp_0 and vm_0

Well wait though, don't you want 0,0,0 to be the center of gravity......

Sun isn't always the exact center is it? It's shifted a bit usually based on planetary mass?

@HelkaHomba Huh what? The Sixty Symbols video? I'm actually not subscribed to that one.

@AshwinGupta we're assuming that's negligible

restricted 3 body problem

@Maltysen I see...

Anonymous

5:13 AM

The sun, planet, and moon have masses M_s, M_p, and M_m

@Downgoat can you link me to ur prj really fast I wanna see something

@AshwinGupta vihan.org/p/SSS

A: Sandbox for Proposed Challenges

gracias

New Sandboxed Posts

0

Molecular Geometry With all the many questions on this site, I am very surprised we don't have that many about quantum mechanics! I mean, such an amazing subject, and we really have little to none about that! Maybe only 5 which I could find by searching, and the most recent one of those is from ...

Anonymous

The acceleration felt by the planet is G*(M_s/xp_0^2 + M_m/(xp_0-xm_0)^2), and the acceleration felt by the moon is the same but replace xp with xm in the first part of the sum

Anonymous

5:16 AM

Your position at time t is x_0 + v_0*t + 1/2*a*t*t

@Mego that's what he's doing, except his time steps are too big so everythings inaccurate

Anonymous

Well make the time steps smaller

Geobits

>_<

Anonymous

Rather than 1s/s, do 1ms/s

Anonymous

Or do 1s/s but do it in ms chunks

5:17 AM

that's what geobits and I have been telling him for the last 2 hours -.-

Anonymous

Oh

Anonymous

@Downgoat Listen to Geobits and Maltysen they're smart

@Mego yes I am

but he can't do that because he has his logic with his rendering

Anonymous

Wait what

Anonymous

5:18 AM

MVC your shit

@AshwinGupta This is cool. By the way, when you get a spacebar press, do stop the event from propagating so the page doesn't scroll down like the normal behavior is.

@AshwinGupta bugfix: you're checking you boundaries based on x,y not left,right,top,bottom

so your spaceship can halfway leave the screen

so I'd do: $y_2=v\frac{y}{\sqrt{x_1^2+y_1^2}}\\x_2=v\frac{x}{\sqrt{x_1^2+y_1^2}}$ to get the new components correct?

@El'endiaStarman ty, its obviously a serious WIP. I've never done JS/HTML before (or much game programming at all) and I started only a few days ago, hopefully I can make it something neat. Not original btw, total Galaga copy, Thanks for the suggestion, I noticed that too, I honestly have no clue how to fix it but I'll add it to my list and look it up l8r lol

@Maltysen yeah I'm lazy aren't I? I'll fix it tmrw

likely during class...

@AshwinGupta return false from your event handler

5:21 AM

@Maltysen meaning?

that stops it from propogating

Oh

one se

sec

Or use event.preventDefault().

um

I tried to return false, not working

I'll try Starman's

That reminds me, @CᴏɴᴏʀO'Bʀɪᴇɴ, would you mind transferring ownership of the Spacewar! KotH GitHub repository to me?

5:24 AM

@El'endiaStarman Lol, spacewar is probably not the best original name for a game like mines huh?

@El'endiaStarman should I just throw that in both of my event handlers (keyup and keydown)?

~~oops I think I did it wrong, it broke xD~~ nvm, typoed, got it TY

@El'endiaStarman & @Maltysen fixed it TY guys. Fixed Version Here. GN now, will be back at it tomorrow

@Downgoat one of my moons became a planet -.-

nice

so I implemented what @Mego said and it doesn't work :/

let sm_axis = /*some value*/;
let v_op = Math.pow(pl.velocityX, 2) + Math.pow(pl.velocityY, 2);
let v_mag = Math.sqrt(
    f * pl.mass * (
        2 / Math.sqrt(
            Math.pow(pl.x - (pl.x + sm_axis), 2)
        ) - 1 / sm_axis
    )
);

moon.velocityX = v_mag * Math.cos(pl.velocityX / v_op);
moon.velocityY = v_mag * Math.sin(pl.velocityY / v_op);

@Maltysen do you know why the above isn't working ^

pls

Anonymous

5:47 AM

Well part of the problem is that you're making the moon's direction be the same as the planet's direction

@Downgoat did you take the normal of the vector?

ninja'd

oh, yeah, fixed

Anonymous

The other problem is that you're trying to compute a unit vector using d^2 instead of d

@Mego what do you mean by that?

^^ @Downgoat what have I been telling you for the last 2 hours?

(╯°□°）╯︵ ┻━┻

5:48 AM

@Maltysen I fixed that alreadu

@Downgoat he means's you have to take the perpendicular of the slope

so -v_op/velocityx

Anonymous

Math.pow(pl.x - (pl.x + sm_axis), 2)

Anonymous

...so Math.pow(sm_axis,2)

and then the lovely Math.sqrt() around that

@Maltysen on the last two lines?

5:50 AM

@Downgoat if I'm reading your code correctly

Anonymous

v_op needs to be Math.sqrt'd

>NaN
NaN
NaN

Anonymous

So let v_op = Math.sqrt(...)

dammit JavaScript

Math.sqrt(Math.pow(pl.x - (pl.x + sm_axis), 2) == sm_axis ??

Anonymous

5:52 AM

Euclidean norm is sqrt(x*x+y*y)

let sm_axis = (a[3].value + "00") / 1e4;
let v_op = Math.sqrt(Math.pow(pl.velocityX, 2) + Math.pow(pl.velocityY, 2));
let v_mag = Math.sqrt(
    f * pl.mass * (
        2 / Math.sqrt(Math.pow(pl.x + sm_axis, 2)) - 1 / sm_axis
    )
);
// Moon vectors components
moon.velocityX = v_mag * Math.cos(-v_op / pl.velocityX);
moon.velocityY = v_mag * Math.sin(-v_op / pl.velocityY);

Anonymous

You're doing r = Math.sqrt(Math.pow(pl.x + sm_axis, 2)) right now, which isn't correct

Anonymous

r is supposed to be the distance between the moon and the planet

Anonymous

So r = Math.sqrt(Math.pow(pl.x-moon.x,2) + Math.pow(pl.y-moon.y,2))

sm_axis is the distance from the planet to the moon so I can just use v_mag = sqrt(f * primary.mass * (1 / sm_axis)) correct?

Anonymous

5:55 AM

Of course, all of this is pointless, because it doesn't factor in the star's gravity for the moon's revolution