Comprehensive Wishlist

[L`zard]

ORIGINAL: Veers

*shakes head*
That's so wrong on so many levels....[:-]


and yet still funny....[:D]

Yeah, I know, but someone has to say these things!

'sides, it's true!

[ColinWright]

ORIGINAL: Veers

*shakes head*
That's so wrong on so many levels....[:-]


and yet still funny....[:D]

Don't fret.

Those automated internet scanners at the FBI have probably already picked his post out. Lizard could already be regretting his post.

[Veers]

ORIGINAL: ColinWright

ORIGINAL: Veers

*shakes head*
That's so wrong on so many levels....[:-]


and yet still funny....[:D]

Don't fret.

Those automated internet scanners at the FBI have probably already picked his post out. Lizard could already be regretting his post.

HAHAHA...Yeah.

GOOD LUCK, L'ZARD! [:D]

[Curtis Lemay]

ORIGINAL: MarGol
I was merely thinking of an alternative to plane maps to be implemented in the game engine itself. By having a spheroid grid mapping the earth's surface, map projection will not be an issue. The hard bit is setting up the hex relations (which hexes are considered adjacent to a particular hex). This is non-trivial, but is done on a regular basis when generating grids e.g. for fluid-dynamical computations (I believe triangular grids are often used, which could be useful here since hex centres form a grid of equilateral triangles).

For example, if we arbitrarily choose the North Pole to be hex (0,0), it would be surrounded by hexes (0,1) through (5,1), and so on going southwards. Down at the equator, hex (0,100) would be adjacent to (399,100) or so @ 100 km/hex. Finally, the South Pole would be hex (0,200) or something, surrounded by (0,199) through (5,199). Scrolling around in your map window would be like Google Earth, but since you have such a small patch of the map visible, the curvature would not be a big issue (at the normal zoom level for game play, my [admittedly low] screen resolution usually lets me see 20 hexes width at a time, which would be 2000 km @ 100 km/hex). Hence, the map window could probably look and feel as usual (however, I guess the minimap would be something else!).

I think it's far more complicated than what you've indicated. The number of hexes per latitude row would have to be specifically correct to get a sphere. Otherwise, you could end up with, for example, two cones butted together instead - or any other variation on a top. I'm not sure that hexagonals fit together on a sphere at any hex count, much less an arbitrary one. I also don't think the hexes would be displayed in the nice neat rows and columns we have hard-coded in now. They would have more and more curvature to them the closer to the poles one got - ending in a circle right at the pole. Plus, the hexagonals used now aren't regular - they have extra width. Display operations would have to be completely re-coded with new graphics, even if it proves doable. For what? What topic requires the entire world including the poles?

I think a more achievable goal would be to model a cylinder. Then, when you right-scrolled hex-column 399 onto the map, hex-column 0 would appear next, etc. That should be good enough for someone wanting to model both the ETO & PTO on the same map.

[Martin_Goliath]

ORIGINAL: Curtis Lemay

I think it's far more complicated than what you've indicated. The number of hexes per latitude row would have to be specifically correct to get a sphere. Otherwise, you could end up with, for example, two cones butted together instead - or any other variation on a top. I'm not sure that hexagonals fit together on a sphere at any hex count, much less an arbitrary one. I also don't think the hexes would be displayed in the nice neat rows and columns we have hard-coded in now. They would have more and more curvature to them the closer to the poles one got - ending in a circle right at the pole. Plus, the hexagonals used now aren't regular - they have extra width.

In order to cover the earth's surface (and not, e.g., a cone), the "hexes" will not be perfect hexagons, but need to be slightly deformed. However, once you have worked out the grid (i.e. neighbour relations) for these deformed hexes, you can always view a portion of the map using perfect hexagons. This would be a good approximation, as long as the chunk you are viewing is small compared to the size of earth. Zooming out would mess things up sooner or later!

ORIGINAL: Curtis Lemay
For what? What topic requires the entire world including the poles?

SAC sortie over the polar cap? Shades of Dr. Strangelove[:D]
Yes, that's the main objection. With limited programming resources and an already impressive wish list, this does probably not qualify. Your suggestion with a cylindrical grid might be worth considering, though.

[ColinWright]

ORIGINAL: Curtis Lemay


For what? What topic requires the entire world including the poles?

Mars Attacks. Any other strange questions?

Anyway, a global map would be cool. How practical or necessary it would be is another matter.

[ralphtricky]

http://www.neubert.net/BUCKmins.html

Buckyballs and Fullerenes.

Honycombs looks like a good start. I suspect that if we ignore the area right around the pole, that might be good enough.

I think it limits it to specific radii, though. It probably makes more sense to just have different number of hexes in each row, but that would make rotating difficult, I'm not sure how you could rotate without somehow adding an extra row of offsets somehow, maybe in the oceans. <scratches head> It might look strange though on the seam. It's a 2D projection which makes things much more difficult.

I think that a cylnder may end up having to be good enough if it ever comes to that. I'm not sure who would play such a monster, though.

Ralph

[rhinobones]

ORIGINAL: ralphtrick
Buckyballs and Fullerenes.

Actually Buckyballs could be an excellent start for a spherical TOAW map model.

The cylinder idea makes no sense. It just reduces four map sides to two map sides. Would be much better if TOAW progressed to the spherical model where there are no artificial map sides.

Regards, RhinoBones

[ColinWright]

ORIGINAL: ralphtrick

http://www.neubert.net/BUCKmins.html

Buckyballs and Fullerenes.

Honycombs looks like a good start. I suspect that if we ignore the area right around the pole, that might be good enough.

I think it limits it to specific radii, though. It probably makes more sense to just have different number of hexes in each row, but that would make rotating difficult, I'm not sure how you could rotate without somehow adding an extra row of offsets somehow, maybe in the oceans. <scratches head> It might look strange though on the seam. It's a 2D projection which makes things much more difficult.

I think that a cylnder may end up having to be good enough if it ever comes to that. I'm not sure who would play such a monster, though.

Ralph

All of World War Two at squad level! Is it going to work to have 50 meter hexes and 250,000 fifteen minute turns?

...we'll need more unit slots, too.

[L`zard]

All of World War Two at squad level! Is it going to work to have 50 meter hexes and 250,000 fifteen minute turns?

...we'll need more unit slots, too.

Not a problem!

All that's needed is cloning Ralph and the Toads, putting them in a 'perfect Arcology', and waiting until they break out and take over the world, eh?

Play it @ 1:1, lol!

[;)]

[Martin_Goliath]

ORIGINAL: ralphtrick
http://www.neubert.net/BUCKmins.html

Buckyballs and Fullerenes.

If I understood it correctly, the buckyball and other fullerene structures all have the problem that a few pentagons are mixed in with the hexagons. The last item (from Deutscher Wetterdienst) was amazing, though: a grid covering the earth @ 60 km, consisting of 163830 hexagons and 12 pentagons!

ORIGINAL: ralphtrick
It probably makes more sense to just have different number of hexes in each row, but that would make rotating difficult, I'm not sure how you could rotate without somehow adding an extra row of offsets somehow, maybe in the oceans. <scratches head> It might look strange though on the seam. It's a 2D projection which makes things much more difficult.

If one can construct a grid consisting of (slightly distorted) hexagons (as opposed to the prefect hexagons + a few pentagons), I guess we are almost there. Maps (and screens) are flat, but the grid specifies a curved surface in 3D. It does not need to have hexes in neat rows at all - I suppose that, as far as the game engine is concerned, the grid is just a specification of which are the six neighbours of a given hex.

The 2D projection does not need to show any distortions of the grid: starting from a hex in the center of view, draw it as a perfect hexagon; draw its neighbours as prefect hexagons; and so on until you have covered the whole view. Provided that this view is small compared to the earth, map distortions due to drawing "hexes" as perfect hexagons will be mild. (Zooming out, they will be not-so-mild!)

In the end, it all boils down to whether it is at all possible to cover a sphere with such a grid (each point having exactly six neighbours). I asked a couple of guys at work doing grid generation for "number crunching" applications, but off the top of their heads, they could not give a straight answer - "Interesting problem" and "might work provided that the number of grid points is large enough" was all I got [8|].

(Sorry for not letting go of this - I sort of got hooked on the problem as such [:)])

[BillLottJr]

[/quote]

Don't fret.

Those automated internet scanners at the FBI have probably already picked his post out. Lizard could already be regretting his post.

[/quote]


HAHAHA...Yeah.

GOOD LUCK, L'ZARD! [:D]

[/quote]

They'll still let him playtest my scenario from Git'mo, right?

[ralphtricky]

ORIGINAL: MarGol
(Sorry for not letting go of this - I sort of got hooked on the problem as such [:)])

No problem. I did too, that's where the buckyballs came up. And the headache<g>.

I wonder if it would be possible to link into Google Earth too as an option, or if they'd want a fee<g>.

It would look funny for the early scenarios, but great for the modern ones.

Ralph

[ralphtricky]

ORIGINAL: MarGol

ORIGINAL: ralphtrick
http://www.neubert.net/BUCKmins.html

Buckyballs and Fullerenes.

If I understood it correctly, the buckyball and other fullerene structures all have the problem that a few pentagons are mixed in with the hexagons. The last item (from Deutscher Wetterdienst) was amazing, though: a grid covering the earth @ 60 km, consisting of 163830 hexagons and 12 pentagons!

ORIGINAL: ralphtrick
It probably makes more sense to just have different number of hexes in each row, but that would make rotating difficult, I'm not sure how you could rotate without somehow adding an extra row of offsets somehow, maybe in the oceans. <scratches head> It might look strange though on the seam. It's a 2D projection which makes things much more difficult.

If one can construct a grid consisting of (slightly distorted) hexagons (as opposed to the prefect hexagons + a few pentagons), I guess we are almost there. Maps (and screens) are flat, but the grid specifies a curved surface in 3D. It does not need to have hexes in neat rows at all - I suppose that, as far as the game engine is concerned, the grid is just a specification of which are the six neighbours of a given hex.

Actually the game treats the hexes as cartesian coordiantes. Trying to project them one a sphere may introduce some non-euclidean geometry, although I may be able to ignore those effects at the scales that we're talking about. My concern is that at the poles, There are fewer hexes in a row, or the hexes need to be smaller.

Apologies if I used the wrong terms, my math was a while ago.

The 2D projection does not need to show any distortions of the grid: starting from a hex in the center of view, draw it as a perfect hexagon; draw its neighbours as prefect hexagons; and so on until you have covered the whole view. Provided that this view is small compared to the earth, map distortions due to drawing "hexes" as perfect hexagons will be mild. (Zooming out, they will be not-so-mild!)

I think if I can find the time, I'm going to have to try it out and see if it works.

In the end, it all boils down to whether it is at all possible to cover a sphere with such a grid (each point having exactly six neighbours). I asked a couple of guys at work doing grid generation for "number crunching" applications, but off the top of their heads, they could not give a straight answer - "Interesting problem" and "might work provided that the number of grid points is large enough" was all I got [8|].

(Sorry for not letting go of this - I sort of got hooked on the problem as such [:)])

It sounds like it's possible, but there may be a gap at the pole where pentagons are needed. If I ignore the poles, then it might work out.

[Legun]

ORIGINAL: ralphtrick
I think if I can find the time, I'm going to have to try it out and see if it works.

Please, Ralph, NOOOOOO [&o]! There is so many important problems - changes improving each of scenarios. The Poles fight in many battles - the poles aren't important in any of TOAW campaigns [;)].

[Curtis Lemay]

ORIGINAL: MarGol
The 2D projection does not need to show any distortions of the grid: starting from a hex in the center of view, draw it as a perfect hexagon; draw its neighbours as prefect hexagons; and so on until you have covered the whole view. Provided that this view is small compared to the earth, map distortions due to drawing "hexes" as perfect hexagons will be mild. (Zooming out, they will be not-so-mild!)

This is incorrect. On a spherical model, each latitude row will have fewer and fewer hexes as one gets closer and closer to the pole. No row has the same number of hexes as its neighbors. The pole has one hex, which would stretch the entire length of the map. That has consequences when displaying it on a flat projection. Note that the flat projection has to show the relationship between units and their movement paths over the map for combat and movement purposes. So it has to reflect the true relationship between hexes. It simply can't do that with the hex grid we have hard coded now.

One option would be to dispense with hexes altogether and switch to rectangles in a brick pattern. Then, the flat projection could show greater and greater length to the rectangles the closer to the poles one got. There might be unintended consequences to that that I can't envision now, though.

The other option would be to dispense with basing the flat projection on the above cylindrical transformation and go with a planar projection over whatever was the center hex. But, in that case, as I've said before, the hexes aren't in neat rows and columns like now - they will have all sorts of patterns. It will be incredibly complex.

Either way, a complete revision of how the map is displayed, including vast graphical work, would be required.

[Curtis Lemay]

ORIGINAL: MarGol
The last item (from Deutscher Wetterdienst) was amazing, though: a grid covering the earth @ 60 km, consisting of 163830 hexagons and 12 pentagons!

Of assorted sizes.

[Martin_Goliath]

ORIGINAL: Curtis Lemay

One option would be to dispense with hexes altogether and switch to rectangles in a brick pattern. Then, the flat projection could show greater and greater length to the rectangles the closer to the poles one got. There might be unintended consequences to that that I can't envision now, though.

What, wargaming without hexes?! [:D]
As you say, rectangles will distort areas away from the equator, like a World map in Mercator projection where Greenland is bigger than Africa.

ORIGINAL: Curtis Lemay
The other option would be to dispense with basing the flat projection on the above cylindrical transformation and go with a planar projection over whatever was the center hex. But, in that case, as I've said before, the hexes aren't in neat rows and columns like now - they will have all sorts of patterns. It will be incredibly complex.

Yes, that's probably what my idea would lead to.

ORIGINAL: Curtis Lemay
Either way, a complete revision of how the map is displayed, including vast graphical work, would be required.

... and, if no other advantages can be seen, it's much effort for little gain

[Martin_Goliath]

ORIGINAL: Curtis Lemay

ORIGINAL: MarGol
The last item (from Deutscher Wetterdienst) was amazing, though: a grid covering the earth @ 60 km, consisting of 163830 hexagons and 12 pentagons!

Of assorted sizes.

Surprisingly, the hexes are all of the same size, see this link.

[Curtis Lemay]

ORIGINAL: MarGol

ORIGINAL: Curtis Lemay

ORIGINAL: MarGol
The last item (from Deutscher Wetterdienst) was amazing, though: a grid covering the earth @ 60 km, consisting of 163830 hexagons and 12 pentagons!

Of assorted sizes.

Surprisingly, the hexes are all of the same size, see this link.

I can't find anything on that link. I was relying on Ralph's original link which had this comment:

"... The coordinate system is based on the icosaeder, which is subdivided into triangles, which are of nearly equal size and therefore ..."

Perhaps that translates into hexes of the same size but various shapes, instead. I don't know. I also doubt the thing has them arranged in latitute rows either. Who knows how much it cost them to develop it - and each change in # of hexes per sphere might require the same effort.