A game that involves exploring a sphere - possible on NES?

rainwarrior · Post by **rainwarrior** » Sat Jun 16, 2012 7:53 pm

I didn't mean that King's Quest was spherical (it is not). I just meant that if you have discrete screens without having to make smooth transitions then linking the screens together as a spherical world is not too tricky to pull off.

At its simplest you could have a cube world with 6 screens. 4 of the screens would be the sides and link together intuitively, and the 2 pole screens would link to the top/bottom of the 4 sides from each side.

Since it would be 6 static screens, each of them could be a proper view of the sphere from one axis-aligned direction, and when you walk off the edge of this view's curved grid, you swap screens and end up at the appropriate location on the next view's grid. It would be pretty intuitive to get around on, I think, even if the rotation ends up being weird at the poles.

Actually, if you made four versions of each side (for each 90 degree rotation) you could make every transition entirely intuitive (i.e. exit on the right, appear on the left on the next screen).

RLError · Post by **RLError** » Sat Jun 16, 2012 8:21 pm

That's actuallly similar to the idea I had. It's not perfectly ideal of course, but at least it is something that covers the important aspects. And it would look pretty good in action..

There is also this: http://mathworld.wolfram.com/HauyConstruction.html

Which is kind of cool, but again, it's also somewhat weird. It is a stretch of the imagination to represent a sphere with it. Still...

Also, I had a feeling I was misunderstanding that thing about King's Quest. I looked up some information on that game. I feel like I missed out by not having ever played it because it looks kind of cool.

Bregalad · Post by **Bregalad** » Sun Jun 17, 2012 2:53 am

Shiru wrote:
Bregalad wrote:Press B+Select on the world map in Final Fantasy II after you got the "Ring" item (which is the first mission).
By 'real time sphere mapping' I mean an effect like this (also used as 'magnifying glass' in other demos), not a boring static picture that is certainly not 'real time'

I know ! Final Fantasy II does such an effect if you press the D-Pad buttons, although it's extremely slow. I think it does this by some kind of bi-linear aproximation, of course it doesn't do exact calculations.

And yes the overworld wraps arround like a normal 2D maps, which is not what rainwarrior wanted anyways.

An "overworld" that is on the surface of a globe gets weird because it won't neatly wrap from one edge to the opposite edge like the square overworlds that we are used to, which only could be mapped to a torus in 3D.

A torus ? I never though it could look like that in 3D, but you're probably right. Now that you mention it, a ficional world where the planet would actually be a thorus would be very fun !

Back to the origial question, I think a sphere could be simulated just by changing the wrap arrounds. For example in the real world, if you start from russia, go north, and north again, until you get past the north pole, you end up in Canada, which is not obvious at all if you look a standard map. If it were mapped like a video game you'd end up near south africa !

Another difference is that in the real world, you can go east or west forever, no problem. But you can't go north forever, because once you've reached north pole, there is no northern point any longer - all directions points to south.

Post by **tepples** » Sun Jun 17, 2012 7:00 am

If you're considering making a video game whose setting spans an entire class M planet pole to pole, you have to be able to simulate tropical, desert, temperate, and polar climates. And if you skip doing the earth science research and try to project a typical console RPG patchwork map onto a sphere, it'll likely fall into the uncanny valley. Have you drawn up a map yet?

rainwarrior · Post by **rainwarrior** » Sun Jun 17, 2012 9:30 am

Bregalad wrote:A torus ? I never though it could look like that in 3D, but you're probably right. Now that you mention it, a ficional world where the planet would actually be a thorus would be very fun !

Pretty much every wrapping 2D RPG world is toroidal. If it wraps north to south, left to right, it's the only kind of surface this wrapping can exist on.

...now I really want to see the ending of Final Fantasy IV redone with doughnut planets.

Anyhow, there really isn't a way of smoothly mapping a square surface over a sphere. There are always "seams" where you need to twist or squish something, or immediately change north to east, etc. Wrapping can't make a sphere. Wrapping without flipping can make cylinders and toroids. Wrapping with flipping or rotation can make weird Möbius or Klein Bottle kind of surfaces.

RLError · Post by **RLError** » Sun Jun 17, 2012 10:22 am

tepples wrote:If you're considering making a video game whose setting spans an entire class M planet pole to pole, you have to be able to simulate tropical, desert, temperate, and polar climates. And if you skip doing the earth science research and try to project a typical console RPG patchwork map onto a sphere, it'll likely fall into the uncanny valley. Have you drawn up a map yet?

I thought it would be nice to try to make a representation of planet Earth.

Bregalad · Post by **Bregalad** » Sun Jun 17, 2012 10:35 am

@rainwarrior : I thought a little more aobut it, and unfortunately mapping a 2D wrapping map on a torus is not really possible.

If you consider a torus, you'll have to admit the inside diameter is always smaller than the outside diameter. Therefore, when "unwrapped" to 2D, a torus world would not be completely recangle, but a curved shape.

You could approximate the fact the outside diameter and inside diameter are close enough only if the main diameter is much larger, that would translate in a 2D map that has the shape of a very long rectangle (either vertially or horizontally).

If the 2D map would be a square, or close enough, it'd be impossible to map on a torus because this means the main diameter is equal (or very close to) the torus diameter, and therefore the "hole" in the middle of the torus is completely filled. But then the "inisde" diameter is equal to zero, and this is a contradition.

Sorry if I'm not using the appropriate terms, I hope you get the idea.

Still the idea of a torus shaped planet sounds cool.

rainwarrior · Post by **rainwarrior** » Sun Jun 17, 2012 10:44 am

Yes Bregalad, you're right, there would be a distortion to the shape of the map on a 3D Torus. (I suppose it could be a 4D Torus just fine though.)

What I meant is that it maps smoothly and continuously to a torus. Things may be squashed or stretched, but any local point is connected to the correct points around it and in the correct orientation.