What if a video game did physics simulations by having it so that:
>every entity has a region of influence that determines how something entering it will react and how it will react to the thing that entered
>something entering said region of influence can disappear resulting in the change of a parameter for the entity; and vice versa
>a group of entities could compile their code together
For example, a piece of iron could have code that says that if an electron enters, its electron count will go up, and if it has a high electron count it will attempt to send electrons somewhere with a lower electron count; and a bunch of pieces of iron grouped together don't have to each do this separately, playing pass-the-parcel with the electron, but instead they can just act as one.
Also, does anyone have warm fuzzy pictures of awoo?
I need distraction from the fact that it's freezing here and I won't be any warmer for the next 5 hours.
>>343123226
but why
>>343126024
>>343126120
Better efficiency of simulation.
Anything that converts what would otherwise be a stepped simulation to a single mathematical calculation is extremely good.
Plus, I think I see some potential for procedurally generated GLSL shaders with such a system.
That'd be a fuckhuge resource hog I'm sure
Just look at dwarf fortress and space station 13 which respectively have life story and knowledge generation and real time atmospherics calculation. If you even dare to put those games on an advanced engine you'd fry any CPU faster than crysis can load a tree
>>343126024
>>343126314
>>343126796
>>343126909
>>343127547
It may be efficient on a bigger scale like grouping many things together as mentioned, but that doesn't mean the electrons themselves wouldn't be a problem
Making all those teeny tiny thingies floating around (assuming you're going for fidelity) implies you'd need atoms flying around as well to hold the electrons, so that means instead of processing one bulky thing interacting with the world you're processing trillions of smaller things making the world interact with itself. So while you're saving yourself the trouble of blessing everything with programing magic for physics you're still making something that'll eat away at RAM like mad
>>343126796
THAT IS NOT HOW DOG LEGS BEND
>>343128162
Actually the point here is that you aren't simulating the smaller things.
I'm no statistician, but you could probably get it so that if you can input a distribution of any number of objects that doesn't have a massive complexity to describe, you can have it so that all those objects end up where they should be described just as simply as long as you evade three body problems.
Speaking of three body problems, are they proven impossible to solve, or is it still up in the air?
>>343128828
Alright, you lost me
I was already having a hard time following what was going on here, but now my brain is mush. Care to translate that to layman's terms?
>>343126024
>it's freezing here
I want to go where you are
>>343129457
Say that you have a sphere of balloons that is denser in the middle and thin around the edges.
These balloons rise, and the higher they get the slower they rise and the more they are blown by a constant wind.
You could plot the movement of each balloon, but alternatively you could just have it so that the sphere is described by an equation and so is the cloud of balloons.
I'm too lazy and bad at maths to demonstrate the maths behind this, but you could say something like for the first instant (while it's still a sphere) the balloon density is inversely proportional to the distance from the middle of the sphere, and then later while it's the distorted shape you use some weird maths on the first equation to distort the way you measure distance from the middle of the sphere because of those winds and changing speeds of rising.
>>343130360
If you're describing treating game objects like a distribution of deformable points, some engines do stuff like that. There was a cool demo that made the rounds some years ago called OE-Cake. Look it up.
The reason not all games do this is because rigid bodies approximate real-world behavior under a pretty large range of hypothetical conditions, and it's useless to complicate or extend the models when you don't need to. It's impractical for the developers and it's expensive on the hardware.
>>343130360
I think I might be getting a grasp on what's going on here, but that raises another problem: human behaviour
Assuming I got what you meant right, then you're effectively calculating everything going on via equations rather than actual engine influence...on a movie. You can't just use equations to adapt to what a player will do to an object since god knows what a player can cook up with ways to break game physics. You'd have to severely cripple player-world interaction towards objects using that equation so that others will never create unexpected math-proof scenarios or spend literal YEARS creating an engine that'll have all possible interactions down to a bible worth of formulas
At that point you're better off just making predefined animation paths for things to follow rather than even applying physics in real time at all
>>343130983
This isn't for replacing individual rigid bodies though; the idea is that you could replace huge numbers of bodies with these distributions.
Say, for example, a gun that fired a million bullets a second.
You couldn't run that in real time by any normal programming.
Speaking of which, I think Star Ruler 2, even though it had no such weapons in it, did seem to be capable of simulating weapons with obscene numbers of projectiles.
>>343131225
If some things are done with these mathematical formulas and some aren't, then you need to create the combinations of formulas in advance.
If EVERYTHING is done with these formulas, you only need to create the formulas for anything that other things are made of, as long as you've got a way that the formulas can combine and simplify.
>>343126024
>>343131886
>second half
Anon that's not my point. My point is that you'd need an obscene number of formulas simply because of people interacting with the objects in so many weird ways
Take the HL2:EP2 gnome as an example. You can take that jolly motherfucker through the whole game with you so that opens up millions of physics based scenarios of chucklefuckery, and you can't possibly resume that to only a few equations
By the time you're done programming your equations will look like an alphabet soup
>>343131886
So you're talking about associating groups of entities and describing their distribution with a formula, rather than having each entity operate independently?
Wouldn't you have to recalculate the groupings and formulas each time? And you'd still need to account for units becoming isolated and needing to operate as such.
It would be an efficient way of describing groups you don't expect to be disassociated from one another. And rigid bodies solve exactly the problem of having to simulate countless particles individually, by just considering them a singular, homogeneous entity.
You already admitted to being lacking in the mathematics department; I don't think you quite understand the limitations of your idea here. It's clearly half-baked and you haven't held it to ample scrutiny.
>>343132273
A-awoo?
>>343132662
Awoo, eh?
>>343132602
It only gets complicated if you have to handle collisions, e.g. sending bouncy balls through a tube.
If you send a charged particle through a magnetic field, it only takes one equation to handle any field strength, direction, particle mass, and particle charge.
The flight of an object, if you're willing to simplify the aerodynamics, can be calculated based on mass, drag, initial velocity, and gravity.
>>343132648
Hmm, finding formulas for distributions of objects would be a problem.
It's possible to describe a mathematically describable collection of objects with a formula that gets more complicated each time the collection is interacted with, but that can only go on for so long before that formula is just unworkable.
And rigid bodies don't solve the need of having to simulate countless individual objects, there's too many situations where individual objects moving as a solid swarm is no substitute for a fluid one.
>>343132273
awoo~
>>343133737
Are you by chance implying there won't be an insane number of collisions throughout any given game? I think >>343132648 here is right, you gotta examine this way deeper than you have thus far
I give up
>>343134174
Collisions are the problem, but it's really a question of what statistics can and can't do.
>>343126024
You know what will keep you warm? Dancing the Awoo~!
>>343126909
Hey! What's so funny awoo?
>>343135703
Your face!
>>343135820
Awoo! You...!
post more lewd awoo
>>343136013
Who needs awoo when you got meow.
>>343136247
its not the same my man
Don't forget to Wan too!
>>343136284
JUST POST HER CUNT ALREADY
>>343136461
thank you
>>343137194
>>343136708
You people are lewd!
