Map Generator Model

[Gary Thomas]

I forgot to mention what I think is one of the most important benefits of the polygon approach.

Because locations are kept in the form of coordinates (rather than merely listing the square you are in) there is no need for a regular tessellation of the playing area. While I have been assuming that the coordinates are rectangular cartesian coordinates, they need not be.

In particular, they can be lattitude and longitude.

This means that a spherical world map is perfectly possible.

The full world map could be rendered in any desired projection, or viewed as a globe.

The local view would be designed to minimize local distortion.

All movement would be calculated in great circle terms.

Cheers

[Simon Loverix]

I mean what happens if a flood occurs and the polygon is half outside, half inside the range of effect? It would be split. (2 pieces) (Squares don't need to, since the range of effect is already defined in squares. ) And when something else happened in the area, it would split again (3 or 4 pieces). And so on, until all polygons are reduced to the smallest possible size. I just think it would lead to fragmentation ad infinitum, so the map would boil down to a lot of tiny little pieces.. like squares. You could also impose a minimum distance and standard ranges of effects to keep the polygons to a decent and minimum size and shape, but then you would lose some, if not all benefits of polygons.

Also, how do you handle cities? You can't have cities if the population (or whatever) is evenly distributed in a polygon.

Polygons have considerable benefits, but I think the scale of the map must be much smaller to be able to use them properly. And a smaller scale means even more detail than we're planning with now.

[Gary Thomas]

I mean what happens if a flood occurs and the polygon is half outside, half inside the range of effect? It would be split. (2 pieces) (Squares don't need to, since the range of effect is already defined in squares. ) And when something else happened in the area, it would split again (3 or 4 pieces).

The range of effect is the polygon. No splitting.

And so on, until all polygons are reduced to the smallest possible size. I just think it would lead to fragmentation ad infinitum, so the map would boil down to a lot of tiny little pieces.. like squares.

That is what happens now. Squares are the smallest possible size.

Also, how do you handle cities? You can't have cities if the population (or whatever) is evenly distributed in a polygon.

To paraphrase "you can't have cities if the population is evenly distributed in a square (which is a polygon)". Why not?

I did not say that the population is evenly distributed, I said that any uneveness doesn't matter.

In any case, cities are constructs which would sit on top of underlying polygons, (just as they do in squares, which are polygons) with population additional to the underlying population. A consequence of this is that you can have more than one city per 100 km x 100 km area, which doesn't happen with the present system, but which could easily, since squares are polygons).

All of these objections apply equally well (or, in some cases even more so) to squares. It is pointless objecting to polygons on grounds that provide an equal criticism for squares (which I think I might have mentioned before, are polygons).

Polygons have considerable benefits, but I think the scale of the map must be much smaller to be able to use them properly. And a smaller scale means even more detail than we're planning with now.

I cannot for the life of me see how this follows. And, as I think I might have mentioned before, squares are polygons, so we had better not use squares either because they will get more and more subdivided into tinier and tinier squares.

It seems to me that the real issue is being overlooked. It really doesn't matter how the map is split up. The way it works now is that EVERYTHING in a square is effectively at the centre of the square, having no other coordinates. So the map does not consist of squares, it consists of a rectangular grid of points, even though we choose to picture it as squares. It is this point location I want to get rid of. So things wouldn't be "in a polygon" as they are currently "in a square". They would be at specific coordinates. As a consequence, it would take time for a unit to cross a square (or polygon). If a unit is near the border, it can cross into the next area in a short time. If squares were a better system, every atlas would have its map areas divided up into squares. Are they? No, they are polygons.

Cheers

[Lord God Jinnai]

Yes well i hope everyone now knows a square is a polygon (but not ness the opposite).

Anyway there is only 1 small problem i can eventually see for using polygons and that is in terms of movemnt of units and how they'd look (i can only concieved of using 3-D rendered graphics which with tons of them on a screen at once will really slow down the comp, even on some of the new ones out today.

BTW...wonder if Marks been reading this forum and if so what he thinks about your ideas for later versions.

[F_Smith]

FWIW --

I completely agree with Gary.

From the beginning, polygons had been in the plans. Altho not in the sense of 'splitting' and all that . . . more like 'Risk' or the original 'Civilization' board game.

Or like the 'Shogun' computer game.

There's no technical reason not to, altho it makes map creation a little more interesting . . .

The biggest bonus to irregular-shaped MapSquares is a simplification of movement, and the 'strategic' importance that a MapSquare can have. The Map can be much more interesting, and a lot more fun.

[Mark_Everson]

Due to LGJs mention of me... (Thanks for thinking of me ) I'll rear my ugly head here for my monthly curmudgeonly comment.

Originally posted by F_Smith
From the beginning, polygons had been in the plans

Hey 'F': From the beginning polygons had been in the Discussion IMO. I certainly don't recall any plans involving them being agreed to, although there has been fairly strong support for a variety of plans at one time or another.

Just so long as we put it off till demo 50 .

Best to All,

Mark

[Gary Thomas]

Just so long as we put it off till demo 50

I had in mind something more like Version 2.0

Yes well i hope everyone now knows a square is a polygon

Glad you noticed!

Anyway there is only 1 small problem i can eventually see for using polygons and that is in terms of movemnt of units and how they'd look (i can only concieved of using 3-D rendered graphics which with tons of them on a screen at once will really slow down the comp, even on some of the new ones out today.

Don't see why it should be any slower than tons of units on the screen under the present system.

In general (as a coding practice) performance issues are dealt with using two rules:

1. Deal with performance issues later.
2. When later arrives, see rule 1.

However, on rare occasions peformance does need to be addressed. The rule then is worry about it when it happens.

From the beginning, polygons had been in the plans. Altho not in the sense of 'splitting' and all that . . . more like 'Risk' or the original 'Civilization' board game.

They have certainly been in MY plans from the beginning.

I don't know where the idea of splitting polygons crept in - it was never part of my model, and I never heard of anyone assume that we were going to split squares (which are...)

Cheers

[Simon Loverix]

It seems to me that we are discussing several things at once:
1 the shape of the area of 1 coordinate (square or other polygon)
2 the scale and the level of detail, including the number of coordinates in a polygon (which can be a square but doesn't have to)
3 the splitting of polygons (that can be squares, but don't have to)

1 doesn't matter, cosmetic
2 Simple, I think: more detail, more power needed, which isn't necessarily a problem but it could become one.
Do you mean the coordinates in one polygon are former square coordinates, or will the total number increase?
In the first case, I think you will still end up with many tiny polygons (of whatever shape) in the more frequently used areas. In the second case we'll just have to wait how many power is still available.
3 If polygons have only one coordinate this is irrelevant, but if they have more than one it is relevant: What if a meteor falls at Tunguska and entire Siberia gets blasted? It barely affected 2000 km² in RL. Will the polgyon split in an affected one and an unaffected one, or will the entire area be affected? If you want range of effects, you need to split polygons sometimes or often. Otherwise you are treating them as we are now.

[Gary Thomas]

1 the shape of the area of 1 coordinate (square or other polygon)

Here, I think, is the source of the misunderstanding. There is no "1 coordinate". A polygon is specified by a coordinate for each corner, and includes everything inside it. The coordinates are the same as the ones used to specifiy a position on the Earth's surface, that is lattitude and longitude.

2 the scale and the level of detail, including the number of coordinates in a polygon

There are an infinite number of coordinates "in" a polygon, one for every point in it. I am not sure what is meant by "the level of detail".

the splitting of polygons

Doesn't happen, ever.

What if a meteor falls at Tunguska and entire Siberia gets blasted? It barely affected 2000 km² in RL. Will the polgyon split in an affected one and an unaffected one, or will the entire area be affected? If you want range of effects, you need to split polygons sometimes or often. Otherwise you are treating them as we are now.

I would apply the effect to all the polygons that are affected. No problem.

Cheers

[Simon Loverix]

By level of detail I mean the total number of coordinates on the map and how many polygons there will be. And hence my persistent questions on splitting: If you have few polygons, what happens if only a small area is affected in a large polygon? Is the whole polygon affected or only that small area (in which case I expect that area to form a new, separate polygon (=splitting), since it has other properties)?

on micro-terrain: Can it be useful somewhere apart from battles and city views?

[Gary Thomas]

I don't understand what you mean by "number of coordinates on the map". There are an infinite number of coordinates on the map (conditioned only by the accuracy of the floating point number format used, or the lower limit of visibility selected). I suggested that a lower limit of 1 hectare be used. On this basis there are about 53,114,285,714 coordinates on the Earth's surface.

If you have few polygons, what happens if only a small area is affected in a large polygon? Is the whole polygon affected or only that small area

You have small and large polygons, depending on context. What happens when a small area of ANY polygon is affected is EXACTLY what happens when a small area of a square is affected, namely, you don't know that only a small area is affected, you assume the whole entity (square or polygon) is affected. Again, I don't understand how you detect that only part of a square is affected.

on micro-terrain: Can it be useful somewhere apart from battles and city views?

I don't think it will be very useful for city views. It might be useful for planning movement, siting fortresses, bridges or ports, or such things. It will certainly make consistent tactical views of battles possible.

Cheers

[Simon Loverix]

Originally posted by Gary Thomas
about 53,114,285,714 coordinates on the Earth's surface.

this, with several layers of polygons.. is what I mean by more complexity. The polygons COULD (depending on chance rather than the design) evolve too small (and they won't stay the same during the whole game- at least the interesting ones).

You have small and large polygons, depending on context. What happens when a small area of ANY polygon is affected is EXACTLY what happens when a small area of a square is affected, namely, you don't know that only a small area is affected, you assume the whole entity (square or polygon) is affected. Again, I don't understand how you detect that only part of a square is affected.

The problem here is that a square is standard, but depending on the proportion between the effect and the polygon size the effects can work out differently. For example, I might assume a certain effect is negligible in a square, and leave it out. But it can be important if the affected polygon happens to be smaller.
That's the point: squares can't get smaller, so you can forget about small-scale (although drastic) effects.

I don't think it will be very useful for city views. It might be useful for planning movement, siting fortresses, bridges or ports, or such things. It will certainly make consistent tactical views of battles possible.

Master of Magic used it for the city views: it made each city unique in it's layout.

Cheers

So long, goodnight

[Gary Thomas]

this, with several layers of polygons.. is what I mean by more complexity.

There is only one layer of polygons.

The polygons COULD (depending on chance rather than the design) evolve too small (and they won't stay the same during the whole game- at least the interesting ones).

Polygons NEVER, EVER, EVER, EVER get smaller, split, evolve smaller, or change their size in any way at all. This must be the tenth time I have said this.

That's the point: squares can't get smaller, so you can forget about small-scale (although drastic) effects.

Polygons NEVER, EVER, EVER, EVER get smaller, split, evolve smaller, or change their size in any way at all. This must be the eleventh time I have said this.

The problem here is that a square is standard,

Why is that a problem, or even relevant?

but depending on the proportion between the effect and the polygon size the effects can work out differently. For example, I might assume a certain effect is negligible in a square, and leave it out.

Apparently, for a square, you decide either to omit the effect, or apply it to the whole square. For a polygon, you decide either to omit the effect, or to apply it to the whole polygon.

Master of Magic used it for the city views: it made each city unique in it's layout.

Might work, would be great if it does.

So long, goodnight

Sleep well. It is 3pm here, so I will stay up a while longer.

Cheers

[Mark_Everson]

You guys need a new thread so Gary has room to say

Polygons NEVER, EVER, EVER, EVER get smaller, split, evolve smaller, or change their size in any way at all.

for the 20th time

Gary, are you up to the honors of starting the new one? The thread changeover is just like the one for Mil IV -> V. But one thing I left out was to put a link to the new thread up in a web page discussion thread so Dom can add the new thread as the current one for this topic.

[Gary Thomas]

I have started a new thread for Map Generator Model since this thread is full. It is at Map generator model II

I do not feel that the continuing polygon discussion belongs in this thread, so I have started a new Polygon and microterrain thread. Please post any relevant discussion there.

Cheers