Bizzare Map Idea

Vultur · January 11, 2003, 13:08

Since a couple of weeks I've been looking to OpenGL tutorials to learn to use it to code something, and I've seen the spectacular maps that can be realized with it.
So, I've had an idea: why don't have a RTS-games-like map?
I've thought a lot looking other games like WC3, Empire Earth, and Starcraft and I've concluded that an isometric map can be not vinculated to "tiles".
If "tiles" concept is removed by more detailed coordinates, so any unit can move and attack freely on the map, with the only limitation of his moving/attacking radius. Also resources will be gathered only in a certain radius from the base.
I know this idea can sound strange, but seen the possibilities of OpenGL, I thought that can be done. See here

DeathByTheSword · January 11, 2003, 13:32

if possible it would be great we would remove tiles out of turnbased game...that is great distance to travel could be calculated in metres per terrain type

Kuciwalker · January 11, 2003, 15:03

First, it would be AT LEAST kilometers per type.

Second, RTS games actually do use a tile system. That's why the units move in the funny directions. They merely spend time "in transit" from one tile to another.

What you really are thinking about is a coordinate system. This would in fact work quite well.

Jeremy Buloch · January 11, 2003, 17:05

I think a coordinate or RTS type system would work perfectly. It would be much easier to move and you wont wasted time trying to move diagnaly and going through alot of tiles. I support the use of this system.
-J.B.-

Leland · January 11, 2003, 21:58

From what I've gathered in bits and pieces in Blake's posts, it's been the idea for some time now to use (rectangular?) tiles for terrain, and true coordinate system for unit movement... yes?

Vultur · January 12, 2003, 12:47

Quote:

Originally posted by Leland
From what I've gathered in bits and pieces in Blake's posts, it's been the idea for some time now to use (rectangular?) tiles for terrain, and true coordinate system for unit movement... yes?

Not necessarly. With a non-tiled map it's possible to have terrain of all shapes, the only thing we have to decide is how much precise the map mut be done.
With a file that contains altitudes (i.e. a .raw file) we can have a similar thing, that contain altitudes, type of terrain, enanchements, etc.
I suppose there is a tons of graphic engines well designed, we must only found the right one.

Leland · January 12, 2003, 13:34

What exactly do you mean by "non-tiled map"? Some sort of polygonal thingamajig?

Kuciwalker · January 12, 2003, 14:17

JB - read my post.

Leland - you can use a coordinate system, just like in the real world. A unit is located at particular coordinates on the map. Terrain features occupy a certain area on the map.

Jeremy Buloch · January 12, 2003, 16:50

Quote:

JB - read my post.

What are you talking about? Please be more specific. Which post do you want me to read?
-J.B.-

Leland · January 12, 2003, 22:42

Quote:

Originally posted by skywalker
Leland - you can use a coordinate system, just like in the real world. A unit is located at particular coordinates on the map. Terrain features occupy a certain area on the map.

Yes, I can imagine how units and bases and other point-based objects can be positioned on the map freely (and this is not a new idea), but what I am not so sure about is how to represent areas like forests or oceans or mountain ranges... they'd have to be polygons of some sort, wouldn't they?

targon · January 13, 2003, 03:45

Any map "features" (forests, oceans etc.) may be represented either by some set of polygons/curves/fractals(?) projected on the map heightfield or like "terrain type" field associated with your .raw map grid points.

Vultur · January 13, 2003, 04:29

Quote:

What exactly do you mean by "non-tiled map"? Some sort of polygonal thingamajig?

Means that any polygon used to draw the map DON'T affect base, units etc behavior. The "coordinate system" must be enough "precise" to make a relistic movement, like all RTS maps.
A map like WC3 or E.E. would be the state of the art, once found the way to implement it.

Quote:

Any map "features" (forests, oceans etc.) may be represented either by some set of polygons/curves/fractals(?) projected on the map heightfield or like "terrain type" field associated with your .raw map grid points.

You've got the point

Leland · January 13, 2003, 05:48

Thanks for the clarification.

targon · January 13, 2003, 05:58

Quote:

Originally posted by Vultur

Means that any polygon used to draw the map DON'T affect base, units etc behavior.

IMHO, they DO affect unit movement etc., but indirectly: the map has some topographic data which are used both for drawing textures/features/etc and for calculating unit movement, LOS, path-finding etc.

Vultur · January 13, 2003, 12:16

Quote:

IMHO, they DO affect unit movement etc., but indirectly: the map has some topographic data which are used both for drawing textures/features/etc and for calculating unit movement, LOS, path-finding etc.

Obvious, terrain type will affect unit movement, I thought was clear...

targon · January 14, 2003, 03:10

During long history, games were impaired by so-called "attribute buffer" approach, hence the tiles everywhere. Nowdays, that thing is obsolete, IMO, and we may in fact use more general means (like polygonal "features", in terms of Xconq). So screen polygons don't affect anything, they're eye candy while mappolygons have a lot of game-related properties, like unit speed modifier etc. I think some of that features aren't for client side either, due to possible cheating.
BTW, Vultur, how you imagine infamous Fog of War technique applied to your "bizzare map" (natural map, inf fact)?

Blake · January 14, 2003, 06:06

Thats not so bizzare. In fact it's been my intention for a long time to not have unit movement constrained to tile->tile jumps and have the map rendered (basically) as a heightmap.

The coolest way is just to have a function like:
float get_height(float x, float y);
that returns the height at any point on the map.
There would actually be an internal heightmap (raw map data) to specify the general shape of the terrain, this map might have a resolution of 40*40 or something quite small like that.

Then, in the "get_height" function you average the heights of the 4 nearest tiles, and also add some variation from a "detail" function, that could use some sort of randomness or get it's value from another tilable heightmap that might be tiled 10 times over the main map (basically the detail is what targon said a few posts above)

The cool thing about using such a get_height function is you can zoom in or out as much as you want, and the map will be more or less detailed. You do ofcourse need a rendering engine sophisticated enough to work out the polygons that end up on the screen.

Another thing is cities would just be built at a location on the map, giving as absolute co-ordinates rather than tile co-ordinates, armies would move in a simialler way, in fact it would even be possible to eliminate "map polygons" altogether, using a simialler "get_terrain" function that works out what the exact terrain is on the fly. Whether or not that would be a desirable way to do things would be a different matter, because in general you need to divide the map into some sort of polygons/sectors in order to do path finding.
What will probably happen is the raw map data will be used to create map sectors.

Btw there is no particular reason to have Fog of War on planets, as they will be orbited by ships or spacestations capable of making detailed surface scans, while in some cases it might be slightly more realistic to have it (blockaded world?) not needing to store visibility data is a good thing (Remembering each players fog of war is different)

targon · January 14, 2003, 06:20

float get_height(x, y) =
1) Perlin noise
2) Fractal noise (not Perlin's)
3) Riged multi-fractal
4) "Plasma" (smoothed)
5) Subdivide&Displace? (tooo iterative)

Leland · January 14, 2003, 06:38

6) NURBS surfaces?

Vultur · January 14, 2003, 08:02

Quote:

BTW, Vultur, how you imagine infamous Fog of War technique applied to your "bizzare map" (natural map, inf fact)?

Like RTS games, obvious: if we had a RTS game-like map, I see no reason to have strange/hybrid unit behaviour.
You can immagine a orthogonal projection of the map "seen from the sky" to calculate things like movement/support/view (FoW) radius that look like circles, indipendent from the FINAL look of the 3D map (isometric or "prospective")
The same idea can be applied also for a "space map", if we decide to implement one. I'm currently looking for a good image of the Milk Way Galaxy to use it.

Quote:

Thats not so bizzare.

Thanks

Leland · January 14, 2003, 11:37

About space map... is the game "galaxy" (i.e the star systems available to players. This is not necessarily the whole galaxy of course) going to be based on real milky way, or is it going to be fictional/random one that allows you to put a bunch of habitable solar systems in close proximity to each other?

Vultur · January 14, 2003, 13:00

Quote:

About space map... is the game "galaxy" (i.e the star systems available to players. This is not necessarily the whole galaxy of course) going to be based on real milky way, or is it going to be fictional/random one that allows you to put a bunch of habitable solar systems in close proximity to each other?

It's just to have a nice reference, but for obvious reasons the galaxy will be fictional.
Note: not the whole galaxy is colonizable, and this reduce the playable area".
The image, from a science magazine, shows where Earth-like planet are in a galaxy Milk Way-like. (Section in violet)
Near to the center there are too many radiations, farther too few metals. It's possible to allow the colonization in planet in this state, but will be more difficult/dangerous/...

Jeremy Buloch · January 14, 2003, 20:37

Sounds good I support this idea of having a part of the galaxy that is habitable.

-J.B.-

Straybow · January 15, 2003, 01:37

Astronomers keep finding there are more exceptions when it comes to metals concentrations in galaxies. Pretty soon the "rules" will be the exceptions.

Seriously, though, the scale of a whole galaxy dictates an entirely different kind of game than what we've been considering for StelPol.

Leland · January 15, 2003, 02:42

If I recall correctly, Milky Way is something like 70,000 light years in diameter. Having a "sandbox" of 200-1000 ly across would probably be big enough for the game purposes, yet it'd be just a small dot on the map of the whole galaxy. It would also not be too much of a stretch to just handwave away the quality and quantity of colonizable solar systems in that area.

The thing is, how to limit the colonizable/playable area? What prevents the player from just steering his ship in straight line until it reaches whatever place in the galaxy he wants to be in? What if he wants to go back to earth for example, what happens?

Kuciwalker · January 15, 2003, 19:44

Here's one explanation: the Earth no longer exists. Do to the catastrophic increase of mass of the population of America, the Earth collapsed into a black hole :P

Jeremy Buloch · January 15, 2003, 20:39

The "Exodus" stated that the aliens would neutrilize all intelligent life from Earth not destroying Earth. If Gateway has kept it that way then the Earth wouldn't turn into a Black hole, it would stay there.
-J.B.-