Would the Amiga be able to handle Angry Birds?

[Mrs Beanbag]

Raytracing IS physics, it is the physics of light beams. The mathematics of reflections is EXACTLY the same as that of spheres bouncing off flat surfaces. The difference is that raytracing involves potentially millions of rays, but the rays don't interact with each other. A million rays bouncing off planes is one thing. A thousand rays bouncing off a thousand other rays is... a million things.

[Stuckemuck]

That's not correct - your mistake is comparing a picture which has been raytraced (exactly as you described) with a picture that has some objects which changed there position and there shape. In raytracing you need millions of computed bits to get a picture, in a "normal" game you have about 10 to 50 object and compute their position and shape using a given algorithm. This has absolutely nothing to do with raytracing. You really can't compare them. It's like I want to buy a Fiat 500 for my daily way to work and buy a rocket instead. I can't use all facilities of the rocket and I can't use all calculated bits of a raytracing picture.

[Cpt. Hindsight]

I'm asking myself, if there should be an Amiga version of that game. Angry Birds and Call of Duty represent all I hate about modern gaming. Use your f*cking smartphone, if you really want to play that kind of games.

[Stuckemuck]

Thumbs up - these games are really rubbish. But that's not the point. It's just discussing about if the Amiga has enough power for such a task.

Anyway - I'll go and have some fun with Ruff'n'Tumble this evening ;-)

[Steve]

Exactly, its just a hypothetical discussion. I never said Angry Birds should be ported to the Amiga. Muppet.

[Mrs Beanbag]

Originally Posted by Stuckemuck

This has absolutely nothing to do with raytracing.

It doesn't... it has to do with the claim that physics simulation computations don't take any time. Which is nonsense. Raytracing is physics computations. It was just an example of a physics computation that takes a lot of time.

There might be only 50 objects. But that doesn't mean only 50 computations. If every object can collide with any other, that means, naively, 50 times 50 computations. The more objects you have, the more hairy it gets. And if your objects aren't all simple circles, it gets more complex still. Colliding rectangles is simple enough, colliding rectangles that can rotate at any angle, well... go on... you tell us how you would do it. If you have a pile of objects, as is typical on Angry Birds, every single object is involved in one or more collisions every single frame so you had better know what you are doing. Each collision is roughly equivalent, in terms of maths, to ray tracing a single pixel. So, go on. Ray trace 50 pixels in 1/50th of a second.

[Stuckemuck]

I never wrote that the simulation or computation of physical elements and movement doesn't take time. I wrote that in my opinion (got it?) it it possible with a machine like the Amiga. I won't be the guy to state that I could write this code. But I think a coder who is used to the Amiga and its capabilities would be able to write some routines for these physical circumstances. In this sort of game you just simulate some sort of gravity. Ok, there are some objects to move around. But I think with some tables for the movement and gravity it should be possible. What's so damn complicated in this kind of game? There are some falling bricks and some throwing birds in it, what else? What's REALLY important for this kind of game is the graphical presentation. You need someone who is capable of transfering these physics or tables-of-movement into a smooth and natural animation.

[Mrs Beanbag]

I'm not saying it isn't possible on the Amiga. What I'm saying is it isn't as trivial as you suggest.

Gravity, yes, that's easy. It's simply a downward acceleration. But it's not just gravity. It's also a lot of objects bouncing off each other and spinning round. Collision detection between a large number of objects is tricky enough in itself. It's possible in O(n log n) time using interval trees. It isn't possible in linear time (i.e. it's not just a case of looping through each object because there are multiple dependencies to take into account, when any one of those n objects can collide with any other of them). And then when you go to actually calculate the collisions...

Given that you can narrow the problem down to a collision between two objects. Suppose they are round objects. Both of them are moving. You have to calculate the position of both at the point at which they hit each other. Then you have to do some more vector arithmetic to calculate their new directions, after the collision.

For round objects that's actually quite easy, in theory. It involves a small bit of CPU bashing, involving multiplications. But if you want to calculate collisions between a round object and a rectangular object, that's a little trickier. If that rectangular object can be rotated at any angle, that's even tricker still.

Now, suppose we need to calculate the interactions of TWO rectangular objects, that are rotated at arbitrary angles, and also SPINNING at the same time. I'm going to write this in capital letters. THIS IS NOT A SIMPLE PROBLEM.

Read this and then tell me it is a simple problem:
http://www.gamedev.net/page/resource...ollision-r2604
And this just returns whether or not two rotated rectangles overlap - true or false - we've then got to go on to work out in exactly what way they push each other about.

[NovaCoder]

It's actually a question of how accurate you want to make the physics engine, if you reduced the accuracy of the physics engine calculations enough then it would of course be possible but the game wouldn't play correctly because the 'sand-box' physics is what makes the game-play. So even without actually rendering anything, I don't think an 060 would be enough for an 100% accurate version of Angry Birds.

[Mrs Beanbag]

I think we will have to forget "100% accurate". But I think a physics based game is not beyond a stock A1200, depending on how much we are willing to simplify. A game where all moving objects are round, that could be done. Hmm... maybe something like Bust-a-Move, but with gravity, and not restricting the bubbles to a grid...

[Cammy]

I think Puggsy is a good example of a game that uses physics on the Amiga (and Mega Drive), you have to pick up and throw objects at stacks of other objects, causing them to tumble and bounce around. Sure, there's no rotation, but I think an Amiga version of a game like Angry Birds should be mostly bitmap-based anyway, similar to that Mega Drive demo but better. Exile has pretty nice physics too.

[Amiga1992]

Originally Posted by Cammy

I think Puggsy is a good example of a game that uses physics on the Amiga (and Mega Drive)

Yeah I was thinking along the same lines. You don't need to have the objects rotate and whatnot, just move them around with proper physics and give them "hardness" according to material. Should work .

Compromising and making stuff like this for porting is paramount to a proper conversion. "100%" is ridiculous, when the essence of the game is somewhere else., I think.

[mc6809e]

Originally Posted by Mrs Beanbag

Given that you can narrow the problem down to a collision between two objects. Suppose they are round objects. Both of them are moving. You have to calculate the position of both at the point at which they hit each other. Then you have to do some more vector arithmetic to calculate their new directions, after the collision.

For round objects that's actually quite easy, in theory. It involves a small bit of CPU bashing, involving multiplications. But if you want to calculate collisions between a round object and a rectangular object, that's a little trickier. If that rectangular object can be rotated at any angle, that's even tricker still.

Now, suppose we need to calculate the interactions of TWO rectangular objects, that are rotated at arbitrary angles, and also SPINNING at the same time. I'm going to write this in capital letters. THIS IS NOT A SIMPLE PROBLEM.

Read this and then tell me it is a simple problem:
http://www.gamedev.net/page/resource...ollision-r2604
And this just returns whether or not two rotated rectangles overlap - true or false - we've then got to go on to work out in exactly what way they push each other about.

All this makes me wonder if the blitter's collision detection feature would be useful.

Recall that if the result of a blit is all zeros then the flag BZERO remains set to 1.

First render one polygon to a plane. Then render a second to the same plane using D=A*C. No collision if the two polys don't intersect (BZERO remains 1).

[Mrs Beanbag]

Then your mathematical precision is limited to the screen resolution, and you're still left with the problem of working out the exact point of interaction, which you need to update the speed and spin of the objects.

[Stuckemuck]

Originally Posted by Mrs Beanbag

Read this and then tell me it is a simple problem:
http://www.gamedev.net/page/resource...ollision-r2604
And this just returns whether or not two rotated rectangles overlap - true or false - we've then got to go on to work out in exactly what way they push each other about.

I must admit that was a very useful link and you're absolutely right in saying that it's not a trivial task. I'm really glad I don't have to be responsible for a game like this ;-)

[desiv]

[ Show youtube player ]

Nuff said.. ;-)

desiv

[Amiga1992]

Originally Posted by desiv

[ Show youtube player ]

Nuff said.. ;-)

desiv

That's a very poor implementation. I am sure you can do much better and have actual physics (both on Amiga and Megadrive). Just avoid the objects from rotating. OK, it's not totally real, BUT it would look ten times better than this. I think you can also have force links between the objects to have cohesion that you need to break, like in the original, even though you can't place objects on whatever angle you want.

Even the comments mention Puggsy as an example of proper 2D physic behaviour.

[Mrs Beanbag]

What you probably could do without too much extra hassle, is allow two or more circular objects to be connected together by rigid beams, that don't participate in collisions directly. So you could build larger solid objects out of circular objects.

[lordofchaos]

I always liked the way some of the bridges worked on Brian The Lion with the weight thingy, I know it wasn't real physics but it kinda gave the impression it was. Now if that's possible on a 68000 then surely a 68060 could prove interesting. What about using bobs, are they not unlimited or did that "Dragons" demo lie to me

[Mrs Beanbag]

Bobs would be fine if your objects were round or didn't rotate. It's the rectangular/triangular pieces in Angry Birds that would pose a problem, because they can spin round.