Curious if the "hex" will ever be retired.

[Tige]

This isn't necessarily directed at WitP but I'm curious why hexes are still the preferred way of charting locations and terrain.

Wouldn't using Lat/Long charts and Topo maps allieviate alot of projection, plotting and location issues?

-Tige

[rogueusmc]

An acurate Lat-long layout would be simulating 3-D space...alot of processor power would be required for this.

Base to base distances wouldnt be hard because they are fixed. Any changes in course of a TF would require a recalculation in 3-D space.

[ZOOMIE1980]

ORIGINAL: Tige

This isn't necessarily directed at WitP but I'm curious why hexes are still the preferred way of charting locations and terrain.

Wouldn't using Lat/Long charts and Topo maps allieviate alot of projection, plotting and location issues?

-Tige

You're talking about a 3D layout where virtually every pixel is a game object. YOu are talking about a tremendous amount of static data for just a terrain map. And most game objects would occupy many map elements greatly complicating computations for terrain impact. I imagine the abstraction one gets with hexes is so useful as to be around a quite a while yet. Even RTS games use hexes or rectangles for location abstractions.

[Mr.Frag]

You're talking about a 3D layout where virtually every pixel is a game object. YOu are talking about a tremendous amount of static data for just a terrain map. And most game objects would occupy many map elements greatly complicating computations for terrain impact. I imagine the abstraction one gets with hexes is so useful as to be around a quite a while yet. Even RTS games use hexes or rectangles for location abstractions.

I always chuckled a that ... even Harpoon which used vector based math for *everything* could not deal with the fact the world is not flat. Perhaps your game can be the first eh? [:D]

One of the things that always ticked me off ... naval sighting is completely subject to the curvature of the earth. Radar is subject to the curvature of the earth. Even to this day, one takes a SPY-1 system ... it has a maximum range of about 40 miles and that is the best there is made by man.

Then people complain that two old WW2 ships pass by each other unseen [:D]

[Tige]

I'm showing my ignorance here but the 3-D space thing is throwing me a bit. I understand if you attempt to use a full satellite map includling 3-D elevation would be a nightmare at this scale, but, I don't see the difference the cpu calculating a TF moving from one hex to another vs. moving it from one lat/long fix to another. Just as the cpu looks at hex 01,02 and knows the terrain for that hex is jungle can it not know that 02N-167W is jungle?

If I look at a TF, or an island for that matter, no matter what scale the chart is the object will be at the same location.

-Tige

[rogueusmc]

Even to this day, one takes a SPY-1 system ... it has a maximum range of about 40 miles and that is the best there is made by man.

I believe, that with the curvature of the earth, the horizon is only 15 miles away though.

[denisonh]

I do not think it is strictly the movement, but the calculations for "interactions" with associated ranges (planes, other ships, submarines, mines, land objects, etc..).

Having "discrete" places on a map makes it much easier to manage these interactions.

These actions multiply quickly in a game of this magnitude and scale, and would quickly become a serious computational burden for even a higher level processor.

Distances have to be consistent and make sense with respect to the discrete time periods as well.

The problem gets complicated quickly and requires a certain level of abstraction to make it all "work" in a fashion that makes sense and is computationaly feasible.

ORIGINAL: Tige

I'm showing my ignorance here but the 3-D space thing is throwing me a bit. I understand if you attempt to use a full satellite map includling 3-D elevation would be a nightmare at this scale, but, I don't see the difference the cpu calculating a TF moving from one hex to another vs. moving it from one lat/long fix to another. Just as the cpu looks at hex 01,02 and knows the terrain for that hex is jungle can it not know that 02N-167W is jungle?

If I look at a TF, or an island for that matter, no matter what scale the chart is the object will be at the same location.

-Tige

[ZOOMIE1980]

What you are really talking about, I suppose, is just a more fine grained hex or rectangle approach. What resolution would you use? What's a degree of latitude 60 NM? A degree of longitude varies with latitude. How are you going to store the terrain so you know you are in water, on land, in a swamp, lake, mountain, rice paddy, etc??? Didn't Harpoon try this type of thing? But they didn't have ground units to deal with. ?It might be easier for pure navigational purposes but abstracting terrain effects on LCU's and such, it could be a nightmare of data.

[rogueusmc]

ORIGINAL: Tige

I'm showing my ignorance here but the 3-D space thing is throwing me a bit. I understand if you attempt to use a full satellite map includling 3-D elevation would be a nightmare at this scale, but, I don't see the difference the cpu calculating a TF moving from one hex to another vs. moving it from one lat/long fix to another. Just as the cpu looks at hex 01,02 and knows the terrain for that hex is jungle can it not know that 02N-167W is jungle?

If I look at a TF, or an island for that matter, no matter what scale the chart is the object will be at the same location.

-Tige

The earth, with it's Lat-Long coordinate system is a sphere as opposed to a flat 2-D surface. The grid get's smaller the further north you get. and without consistent grid size, it would require alot of processor power.

[Mr.Frag]

I believe, that with the curvature of the earth, the horizon is only 15 miles away though.

Thats why they stick the radar up high on the superstructure ... the bigger the ship, the higher the superstructure, the farther the radar horizon is.

The smaller (lower) the target, the harder it is to get a sighting ...

Best case is BB vs BB (both quite high). Worst case little patrol boats. Couple of miles at best.

[Belisarius]

ORIGINAL: rogueusmc

Even to this day, one takes a SPY-1 system ... it has a maximum range of about 40 miles and that is the best there is made by man.

I believe, that with the curvature of the earth, the horizon is only 15 miles away though.

Depends on your altitude above the sea level too. The BBs' fire control towers a located at the top for a reason... [;)].

Radar works a little bit different too. EM waves (just like ordinary light) can "bounce" between air layers (of different temperatures) and the ocean surface, just like sonar waves bounce between thermal inclines. This can extend their effective range as the bounce allows for waves to travel around the curvature, but I suspect it's very sensitive to current weather conditions.

[DrewMatrix]

Another thing hexes do is to regularize movement. You are at Mandalay if you are in the same hex. You are not at Mandalay if you are one hex away from Mandalay.

Imagine the complication of supply, combat etc if this unit was almost to the town center (say 2 miles, just at the city limits), that unit 27.916 miles and this other unit 62.534 miles away. One doesn't want to have to do a seperate supply calculation for every unit based on how many meters of distance between that unit and the dock is paved, open, jungle and track respectively. You want "one main road and one track hex". You don't want combat to be different in intensity if this unit is 2 miles and that unit 22 miles from the enemy. You want them to be "in contact for ground combat purposes" or "not in contact for ground contact purposes."

[Mr.Frag]

Didn't Harpoon try this type of thing?

Harpoon dealt with the units in 3D space on a flat distorted map. They just don't have the horsepower to deal with the real curvature of the earth. Most games get around this by simply limiting the amount of space depicted to cut down on the rate of error.

WitP covering such a large area of the world really falls into a brand new class for this type of problem. No one else has really had to deal with it or has copped out by limiting the area into chunks where the error is contained into smaller parts.

[ZOOMIE1980]

ORIGINAL: Mr.Frag

You're talking about a 3D layout where virtually every pixel is a game object. YOu are talking about a tremendous amount of static data for just a terrain map. And most game objects would occupy many map elements greatly complicating computations for terrain impact. I imagine the abstraction one gets with hexes is so useful as to be around a quite a while yet. Even RTS games use hexes or rectangles for location abstractions.

I always chuckled a that ... even Harpoon which used vector based math for *everything* could not deal with the fact the world is not flat. Perhaps your game can be the first eh? [:D]

One of the things that always ticked me off ... naval sighting is completely subject to the curvature of the earth. Radar is subject to the curvature of the earth. Even to this day, one takes a SPY-1 system ... it has a maximum range of about 40 miles and that is the best there is made by man.

Then people complain that two old WW2 ships pass by each other unseen [:D]

Yes, pure surface search is a very hit and miss thing. Radar is LOS, which is what 20-30 miles on the ocean? You have to use airborne surface search radar for more range and l don't think they Hawkeyes and Orions in WWII.

And yes, game mapping, where the context is a large area, ala WitP, with the earth curvature is a REAL PIA! Mercator projections are utterly useless for large scale wargames where distance accuracies are important, especially if you want arctic regions represented in your game. This game does good a job of handling this as I've seen to date.

[ZOOMIE1980]

ORIGINAL: Belisarius

ORIGINAL: rogueusmc

Even to this day, one takes a SPY-1 system ... it has a maximum range of about 40 miles and that is the best there is made by man.

I believe, that with the curvature of the earth, the horizon is only 15 miles away though.

Depends on your altitude above the sea level too. The BBs' fire control towers a located at the top for a reason... [;)].

Radar works a little bit different too. EM waves (just like ordinary light) can "bounce" between air layers (of different temperatures) and the ocean surface, just like sonar waves bounce between thermal inclines. This can extend their effective range as the bounce allows for waves to travel around the curvature, but I suspect it's very sensitive to current weather conditions.

That's not reliable enough for use, though. For intent and purpose, radar is LOS. Even lower freq stuff in 200MHz range. If you can get your radar about 200ft above the water line you can get, as Frag says, about 40 miles, max. You have to use airborne surface search for longer ranges and that was technology just beginning to come on line in late WWII. Then you had ground clutter problems and WWII vintage radar was easily jammed and confused by chaff and other tricks.

[waynec]

What you are really talking about, I suppose, is just a more fine grained hex or rectangle approach. What resolution would you use? What's a degree of latitude 60 NM? A degree of longitude varies with latitude.

if i recall correctly, SPI's monster Pacific War game had top section, middle section, and bottom section on the map where the hex "size" was different to compensate as one moved closer to the poles.

[ZOOMIE1980]

ORIGINAL: Mr.Frag

Didn't Harpoon try this type of thing?

Harpoon dealt with the units in 3D space on a flat distorted map. They just don't have the horsepower to deal with the real curvature of the earth. Most games get around this by simply limiting the amount of space depicted to cut down on the rate of error.

WitP covering such a large area of the world really falls into a brand new class for this type of problem. No one else has really had to deal with it or has copped out by limiting the area into chunks where the error is contained into smaller parts.

Yea, that's exacly how most large scale geo scoped games get around it. Break a large geographic area into smaller component maps. Go off the edge of one component and pop onto the edge of the next one. I did that in a little game I wrote back in 1988 on the old Apple IIc for a modern day Pacific theater thing. Had eight maps....and that did not cover China/Burma/India at all!!

And there are those that want the whole world modelled in the next one? How would you even deal with the map in that? ...I want my subs to go under the Arctic Ice cap and hit those Murmansk bound convoys in the Barents Sea......

[ZOOMIE1980]

ORIGINAL: waynec

What you are really talking about, I suppose, is just a more fine grained hex or rectangle approach. What resolution would you use? What's a degree of latitude 60 NM? A degree of longitude varies with latitude.

if i recall correctly, SPI's monster Pacific War game had top section, middle section, and bottom section on the map where the hex "size" was different to compensate as one moved closer to the poles.

Off topic, but I think that's a picture of Clancy, right? Did you know that one of his inspirations for his early novel, Red Storm Rising, was Gary Grigsby's game, North Atlantic '86?

[rogueusmc]

ORIGINAL: Mr.Frag

I believe, that with the curvature of the earth, the horizon is only 15 miles away though.

Thats why they stick the radar up high on the superstructure ... the bigger the ship, the higher the superstructure, the farther the radar horizon is.

The smaller (lower) the target, the harder it is to get a sighting ...

Best case is BB vs BB (both quite high). Worst case little patrol boats. Couple of miles at best.

I just wouldn't thing a alevation of only...what?...300'?...would double the distance to the horizon though.

[Captain Cruft]

Couple of game engines which don't use hexes are the HTTR series and the CM series. AFAIK they use a 2D grid and 3D grid respectively. They are only dealing with small areas of the planet though ...