Island/Continent Based Map?

[kaka]

Not once in my life have I seen a pond right next to a larger body of water,
which is where most ponds on this map are located. That is what we find unnatural.

This is a completely different scale, mind you.
I don't see no ponds 10-20 meters away from the coast.

[Jackard]

i can. look closely

besides, what about tidal pools?

[RaptorJedi]

This is the area on google maps, and you can zoom right in and there's some decent sized ponds 10-20 meters from the larger bodies of water

http://maps.google.com/maps?hl=en&safe= ... a=N&tab=wl

[Laremere]

It's not that small bodies of water don't occur near lager ones. (I live in Minnesota, the land of 10,000 lakes, I've seen it happen) I think the main problem is that smaller bodies of water are occurring only and to a large degree near larger water.

[Jest]

a.k.a it's not doing it naturally.

[Leonon]

Could inland bodies of water be added after the generation we've been shown so far by calculating where water would flow based on terrain height and soil density and adding ponds or lakes in places where water would be forced to flow uphill before continuing to flow downhill?

EDIT: I've noticed that low altitude areas always being bodies of water isn't exactly realistic, would it be possible to make certain areas that are lower than water level not be bodies of water?

[sikgamer]

I think, instead of having a static water level, the local water level should depend on whether the surrounding area is higher in altitude.
For example if there was a small ditch-like area in the mainland, a pool of water would form there. Obviously not in every case should this happen, or the map will be 99% water. But only when the altitude goes up in a certain amount in all directions from a certain point, should a pool form there.

Also, rivers are very important, have them start at random locations at certain high alititudes, then flow downwards in an erosion-obeying structure. Rivers don't constantly snake around irl unless they've been formed artificially like that or they're following a small path surrounded by un-erodable rocks.

[Leonon]

Pools don't always form in ditches though, depending on how far above the water table the depression is and how porous the soil is. I suggested just doing rivers and lakes and pongs because having realistic random ponds would require calculating the in/out flow of every possible pond/lake to see if water could flow in faster than it flows out or evaporates.

One way to skirt around erosion simulation would be to just use a "shortest path" system based on heightmap and possibly soil qualities. Just random seed water start points and have them try to get to sea level using the shortest path with each vertical unit of distance equaling 1.5 or 2 horizontal units of distance (possibly more if this turns out to form too many bodies of water). Each water start point could create a 1 tile wide shallow stream of water, with the water getting deeper as streams converge. I think this would form lakes, ponds, and rivers that curve mildly to fit the land.


For testing though, just seeding water start points and starting full fledged rivers from them may be a good way to see how it all works.

[sikgamer]

Pools don't always form in ditches though, depending on how far above the water table the depression is and how porous the soil is. I suggested just doing rivers and lakes and pongs because having realistic random ponds would require calculating the in/out flow of every possible pond/lake to see if water could flow in faster than it flows out or evaporates.

One way to skirt around erosion simulation would be to just use a "shortest path" system based on heightmap and possibly soil qualities. Just random seed water start points and have them try to get to sea level using the shortest path with each vertical unit of distance equaling 1.5 or 2 horizontal units of distance (possibly more if this turns out to form too many bodies of water). Each water start point could create a 1 tile wide shallow stream of water, with the water getting deeper as streams converge. I think this would form lakes, ponds, and rivers that curve mildly to fit the land.


For testing though, just seeding water start points and starting full fledged rivers from them may be a good way to see how it all works.

I don't think the map needs to be that complicated in terms of water physics, all we need is a system that has large seas splitting continents and some lakes/rivers dotting the mainland. With Peter's latest map, there are only large seas with large stretches of land containing no rivers/lakes at all. However I agree with your second point.
Technically altitude isn't even in the game, we're just walking on endless flat tiles. Mountains are just a blob of a different type of tiles instead of an actual mountain.

[Leonon]

Technically altitude isn't even in the game, we're just walking on endless flat tiles.

It's not in the gameplay, but Peter seemed to say his fractal maps do generate altitude data and generate the water portion of the map from the altitude data.