I'm a hobby programmer and release my games for free to the public. Unfortunatly I lack the OOP design skills a professional would have - perhaps you can help?

Okay lets make a hypothetical set of requirements for a space game

1. The following entities will exist - spaceship, rock, fuel, missile, alienship, sun, black hole
2. The rock can be blasted into smaller rocks
3. The 'fuel' is just like a rock but the space craft can 'tractor beam' it to get more fuel
4. The sun and black hole has a large gravitational field.
5. You die when you hit the sun but you teleport to a random place when you hit the wormhole


Now the proper gravity law will be applied to all objects. Meaning I would like to see spaceship, rock, fuel, missile etc all able to 'slingshot' around the sun/wormhole.

Now at the moment I have the following code (and it's really annoying me!)

- For each spaceship calculate the gravity to each rock, fuel, missile, sun, wormhole, alienship etc and move spaceship
- For each rock calculate the gravity to each spaceship, fuel, missile, sun, workhole, alienship etc and move the rock
- For each missile calulate the gravity... etc etc
- For each sun ..
- For each wormhole
- For each alienship...
etc etc

Now i'm thinking there has to be a easier way to handle the gravity side of things for all these different objects around the place and i'm thinking its a problem in the way ive designed my objects.

How would you design the above entities given then rules i provided to make gravity calulations easier? Aka condense all the 'for each X calculate the gravity of Y' etc

Really I want to be able to have ONE gravity check like so

- For each object that is effected by gravity, calculate to every other object effaced by gravity

Make a GameObject base class (or super class, or whatever they call it in .net), and store all GameObjects in a list should allow for

Wouldn't that work?

So is there a way to group different objects together in a list? And when you need to, access their properties?

I never programmed .NET, but you need to make a list of GameObject references, at least that's how it works in C++ and Java.
As for property access, properties for gravity (probably mass and velocity in the plane) are bes stored in GameObject, so you can access them easily. Otherwise you need a (virtual) method to pull them out of the object.

I agree with Alberth about the double for-loop and creating a GameObject base class for all game objects. Just two things to keep in mind:

1) there's no point in calculating gravity between ships and rocks, as they'll exert close to 0 gravity on each other. You only need to worry about the massive objects like suns and planets.
2) consider keeping references to the massive objects in a separate array massiveObjects. Then you can do the calculations much more easily and avoid having to loop through N² objects each time you calculate gravity. (Where N is the number of game objects you have)


Have you used inheritance / polymorphism before?

Polymorphism basically means you can treat every object that inherits from GameObject as a GameObject, ignoring any other functionality it might have and only using the functionality you know a GameObject has. That means if you keep mass, velocity and position in the GameObject base class, then you can loop through all the ships, rocks, suns and everything else and access the mass, velocity and position without knowing if it's a ship/rock/sun. I'll try explain better if you don't get it

_________________
Whatever the mind can conceive and believe, it can achieve

Unfortunately the game is more of a physics simulation with a game added on top. Each object has a size and density and the mass is calculated from that. It's possible to use the tractor beam to drag a low density "rock" (think of it as a cloud) into another low density rock and have them combine.

It's also possible to shoot high density rocks and have them explode. The total mass of the universe never changes. For example a 10 mass rock may split into a set of 2, 5 and 3 mass rocks. I've actually seen planets form from bunches of rocks collecting together (there is a proceedure that runs that checks to see if rocks have pooled together for a long time and joins them). I've also seen 'ring' planets form with a belt of stuff around them - some unexpected surprises to this dynamic nature.

I do run into two problems though if you could share your thoughts

- The tractor beam at the moment is just a line of fixed length. I would like to make it a 'chain' that drags behind a ship. Aka if you turn i want it to sling shot around you. I've been reading up on spring physics and hope this can do it however if there is a easier way to do this that you can think of please let me know.

- In space things rotate. I have not been able to find any sources (that I can understand) that describe how rotation starts in a object due to gravity. If anyone has any idea how this could be done well (even if its 'fudged') please let me know.



Wow this is exactly what I need to do. I'm aware of inheritance and use it a lot (it seams to save duplicating code) unfortunately I don't have the knowledge to apply polymorphism. I'll read up on what your describing some more.

If you have the time to give a VB.net or C# example I would be most appreciative. Thank you.

Say you have a hierarchy like this:


Then you would have classes looking like this, in C#



Not sure about the tractor beam thing, but a spring sounds like a reasonable way to do that.

I think rotation starts due to gravitational tidal forces. If two objects pass each other, the two nearest sides are most closely attracted, and as they move away, that makes them start spinning. There's a reasonable chance I'm wrong about this, hopefully someone else pitches in to confirm / correct me.

edit: Seems I'm wrong about the tidal forces. It comes down to conservation of angular momentum and the way planets and suns are formed. http://www.physicsforums.com/showthread.php?t=163047

_________________
Whatever the mind can conceive and believe, it can achieve

I was thinking about the rotation thing yesterday. I only came up with two situations:

1. Two objects strike each other. A glancing blow will result in the objects 'rolling' off each other and tumbling along their new course.

2. Transfer of inertia. If an object is 'captured' by the gravity of another object, it will enter orbit around it. Once the captured object becomes part of the parent, the inertia is preserved as rotation of the overall mass.

I have no idea how the maths for the above would work, I just had some ideas and thought I'd throw them in the mix.

I'm still thinking about the tidal idea. I like it and it sounds plausible, but I'm not sure if it would have enough effect to induce rotation.

_________________
10 PRINT "Bad Monkey ";
20 GOTO 10

The spaceships are exerting gravity on the rocks? That doesn't make sense to me... *scratches head*

Let's think logically about space physics - start at the top and work our way down -which object has the most mass?

The black hole, of course.
Then, the sun
then the rocks, assuming they're huge

IMO, spaceships/"fuel"/missiles/ should not exert gravity as they just dont have the mass, and i would argue that 'wormholes' have no mass at all and therefore no gravity.

So if you want a really dynamic and "living" model, you could step through your physics logic in order:

Black hole in the middle, exerting gravity on the sun. sun orbiting black hole
Sun exerting gravity on the rocks, causing them to orbit the sun
rocks exerting gravity on everything else that's smaller.

Assuming theres enough distance between your celestial objects, and your math/physics calculations are sound, what youl end up with is essentially a living-breathing solar system. You know, minus the black hole. Naturally more than one gravitational pull could apply to any given small object, and so youd have to calculate which pull is stronger based on distance/mass of celestial object.. etc.

Sounds complicated but if you work it out logically and code it in a modular fashion, it should work out fine. If i wasn't on my way to bed and half asleep, id write ya some pseudocode. Maybe tomorrow hehe. hope this helps.

_________________
"None are more hopelessly enslaved than those who falsely believe they are free."
"It is no measure of health to be well adjusted to a profoundly sick society."
"Hope is the first step on the road to dissapointment." -Jonah Orion
http://tankzgame.blogspot.com

As I understand it, every object has a gravitational effect. Some things are too small to exert a meaningful influence, but it is still there. When you drop a ball, the Earth's gravity pulls the ball down, but on a minute scale, the ball pulls also the Earth up. However, for the OP's requirement, most small objects' gravity can be ignored.

I would approach this by setting up a grid of vectors, the density of the grid depends on how much work you want to do and how accurate the simulation needs to be. Each object's gravity field can then be applied in turn to the grid until you have a 'map' of the gravity on the whole playfield. When adjusting your object's velocity, you can simply look at the grid to determine the overall gravitional effect at any given point.

_________________
10 PRINT "Bad Monkey ";
20 GOTO 10