Map Mode

[KrazyHorse]

Wow 007...didn't expect anyone to go into so much detail. I agree with your comments on squashing the poles; it's already done on some maps to avoid the ridiculous situation in which the Mercator projection makes Alaska look as big as the rest of the US. I just did a search; the area-accurate map is called the Peters projection. This is about the geekiest discussion I've had in a while . Map projections, solid geometry and graphics design as they relate to video games. Oh lord.

[007]

the Peterson's map is indeed politicaly correct, but it still does not provide a geometrical model for a random planet.
The longer i think about the map the stronger i'm getting convinced that the game doesn't need tiles at ALL.
I consided tiles a rudiment from paper based game that by a miracle got carried on.

[Geoff the Medio]

he's obviously had as many math classes as I have.

nope. Previous discussion on the subject a few years ago. No math class ever talked about icosahedrons. (hedra?)

As for the modelling the sphere by breaking the faces up into smaller triangles, this doesn't work because it requires that the triangles be non-equilateral and not all the same size. If you take an original triangular face, then put 3 new vertices along the edges of the triangle, then push those new vertices out from the centre of the whole sphere (to better model the surface), the new triangle located directly in the middle of the old triangle is bigger than the other triangles around the outside, as well, the 3 corner triangles are no longer equilateral. (hard to explain w/o a diagram... someone with web space, feel free).

Another point of interest: If you ignore the differences in sizes of the triangles, it's possible to construct a grid of tilted quadrilaterals from a sphere modelled out of triangles. There might be some minor problems getting the grid to line up depending on the numbers involved, but basically, if you have a row of triangles, the row above and below is "flipped" upside down compared to it... (this mirroring works in 3 directions). if you take the top corner of the upper row, and connect it, across 2 triangles, to the rightmost corner of the 2nd triangle over in the row, then down to the bottom corner of the triangle below the original, then up and left to the left corner of the triangle 2 to the left from the original, then back up to the top corner of the origianl, you get a quadrilateral that looks pretty much like a tilted civ-style tile... except that it has 4 vertices at its corner, and 2 internal ones (useful for mounthains or trenches or whatever).

I realize that's hard to understand w/o a diagram... but again, no webspace. I made a nice diagram though...

[dainbramaged13]

Originally posted by Provost Harrison
Forget the idea of a globe, that would be complicated and confusing.

RIght on!

Now civ games have always used quadrilateral tiles. I think the principle of the hexagonal square would be better, personally, as I have seen it used to good effect in other games (PG for example) and would even out movement between squares a bit. I personally think it would be a good move, at least to test...

WTF are hexagonal squares?


What do you mean by isometric, and why do you care so much that it would be so?
I dont understand why you think hexagons would be so much better? It creates too many movement problems to be worth it. AOE can use hexagons because they dont define movement, only building placement and such. I played a turn-based war game (some really old ww2 game) with hexagonal squares at my uncles house once, the movement system was incredibly confusing...

To the math geeks on the forums looking for incredibly complex and drastic solutions to the globe problem - FORGET IT!
It's not realistic, and not even WORTH it!

[Darkknight]

Hexagonal that's what we're doing in Humankind. It's a lot more realistic . The distance you move when you move up/down/sideways on the isometric map is something like 1.44 times that when you move diagonal when its a hexagon they're all equal.

[Darkknight]

Yeah! I just read everyone for a 3D globe.
That's easier said then done though. Spent months on it.

[007]

but, but,

if there is no tiels the Globe is easy.



well geoff we were claimed as math freaks so I'll carry on
filling in the face of octahidron with raws of stacked triangulars (like in pool(billiard)) will allow for all the same equalateral triangles.
Want another math problem?
Devide a square into 3 equal parts with straight edge and pensil only (no measurments).

[Geoff the Medio]

To the math geeks on the forums looking for incredibly complex and drastic solutions to the globe problem - FORGET IT!
It's not realistic, and not even WORTH it!

My suggestion was complicated in terms of how to set it up geometrically, but in terms of gameplay, it would be just like current CivII/III, but wrapped nicely around a globe.... which is what really matters, right?

filling in the face of octahidron with raws of stacked triangulars (like in pool(billiard)) will allow for all the same equalateral triangles.

True, but that's not a globe. That's an octahedron with lots of triangles on it. To make a globe, you'd have to push out the triangles on the inside of the face of the octahedron so it would more closely approximate a sphere, which would also distort the triangles as I described.

That's also basically what I suggested before, except I chopped up the rows of triangles to make standard CivII style tilted tiles. (commonly called isometric, I think)

[KrazyHorse]

Originally posted by 007
the Peterson's map is indeed politicaly correct, but it still does not provide a geometrical model for a random planet.

Well, what I meant was that I wish the designers would take into acount the smallness of the area actually in the polar regions, which is what the Peters map does. In the CivII world map, Siberia, Northern Canada and Greenland together took up half the available land area. While these regions are large, their size was exaggerated by the Mercator projection. The bias grows so that it is impossible to even represent the actual North or South poles on such a map, leading to the problem of excess temperate zones.

[To_Serve_Man]

Ick, Hexagons? come on, arent we a little to old to be playing with Hexagons? To move East or West takes twice as long cause i cant travel a striaght line (diagonal up, diagonal down, etc etc). An Octagon (maybe Actagon?) would be better. With 8 sides I could go all directions in a striaght line....

It was mentioned before to make the map a Point system instead of a box system. I support that idea. The point system is kinda like.... Lords of Magic.

[007]

I want to see u draw octagonal map, ok, just 10 tiles.

[marcuspeddle]

Regret my vote a little

I voted for hexagonal because I used to use that kind of graph paper for making maps when I played Advanced Dungeons and Dragons. It was great for paper maps but I can see how it might cause some problems for moving in a computer game. Not only that, what keys would you use for moving on a hexagon map? Well, 7,8,9 for the upper half of the hex and 1,2,3 for the lower half of the hex might work, I suppose.

[Darkknight]

DOH!
You cant draw a octagonal map. Hexagonal is the best you can get I'm afraid, if you want it to be realsitic. Co-ordinate system still uses squares. The best way is to use a globe made of triangles. Each triangle is excactly the same.

[To_Serve_Man]

Hey your right!!! You cant make an octagons hook together! hehe, silly me I didnt even think of that.. I guess we really aren't to old for hexagons... hehe oh well... i still think that a point system would be better... or Civ2 style

[KrazyHorse]

Okay. If you don't mind, check this out. It took forever to draw in Paintbrush. F***.