Anyway to the reason for this message,
I'm helping with the terrain graphics, and I've been working nights to come up iwth ideas for what the terrain MIGHT POSSIBLY look like (Nothing Definite)
If you'd like a look, go to : http://wordsup.homestead.com/clash.html

here is a picture of what a few tiles together would look like.

The whole old art team has unfortunately either resigned or disappeared. Paul Crocker, especially, got a lot done before he left, and I think we can still use much of the previous discussion we had within the old group. Now, onto the new stuff...

Magnus Pettersson is interested in doing work on the tiled graphics for Clash, as well as other art. Welcome Magnus! Hopefully he will stop by this thread and say hi. In the meantime, I have posted some of his initial examples of tiled work he could do, and we have discussed possible graphics formats a bit also by email. His tileset examples are here. You can compare them to the ones done by Paul Crocker that are in the Demo 4 tileset which can be seen on the website map graphics page. Please let us know what you think! I'm going to hold off on my comments for a few days.

The most important thing is that the geography (physical features) has been differentiated from the vegetation on the tile. So instead of just "hills" or "grassland" we have hills with desert, flatlands with evergreen forest, rolling terrain with grassland, etc. This vegetation will change pretty frequently, but the geography probably will not change.

Another big issue is that the game plan calls for turns of as little as a month. As a consequence, the ecology model will model seasonal variation in the vegetation. The art will also have to reflect this seasonal variation.

There are two ways to deal with this:

1) Create a different .gif (or whatever we use) for every combination of geography, vegetation, and season.

2) Define the geography as a "background" and the vegatation as a textured "skin" that is superimposed onto the background. Additionally, the seasonal changes are a "skin" put on the base vegetation art.

I think option 2 would work better. It should give us more flexibility and should be easier to draw for the art team and mod makers. Instead of drawing over a dozen tile graphics for every set of vegetation, they only have to draw one base vegetation texture and a few seasonal textures.

Also, it would make the game look a bit better. If the hills in "hills with forest" and "hills with grassland" have a different shape, then the physical shape of the land would appear to change whenever the vegetation changes. Or even worse, if there was a discrepancy in the seasonal graphics, the shape of the land would appear to change every season. But if only the vegetation skins are changed, the shape of the land will remain constant.

Here is the geography list:

Mountainous
Hilly
Rolling
Flat
Sunken (wetland vegetation only)

The preliminary vegetation list for non-sunken terrain:

Deciduous Forest
Desert (with life)
Grassland
Ice
Plains
Scrubland
Taiga
Tropical Forest
Tundra
Wasteland (no life)

It would be good if the art could represent partial forest cover. For example, an ecological province might be 1/4 taiga and 3/4 plains.

The preliminary sunken terrain vegetation list:

Bog (few plants)
Forested
Lifeless
Swamp (mixed plant cover)

Geology
Vegetation

Geology has a big impact on movement and fortifications. Vegetation has a big impact on economic activities. So the player needs to be informed of both. But drawing all of the combinations seperately is impractical.

I have an idea for doing this quickly:

Draw the geology graphics in grayscale. This simulates borders and shading seen due to the geology. Then, draw textures for all of the vegetation. On the map, put the textures over the geology background. Now you have the vegetation texture with the shading to represent the landscape.

The seasons would simply be a different texture. These are optional, but I think that it would look silly if the land looked the same in winter and summer.

Base underlying landform:
Flat - Fairly even terrain, can be horizontal or sloping continuously. EX: Kansas plains
Hills - Uneven terrain with peaks and valleys, generally no rough terrain. EX: Think of any hill pretty much
Foothills - Between Hills and Mountains. EX: Appalation mountains.
Moutainous - Jagged terrain with sharp peaks that are quite high. EX: Himalayas
Plateaus - 2-layers of flat terrain seperated by a fairly sheer cliff. EX some areas of the continetal shelf

Do you think it is necessary to differentiate hills and foothills? Will they have different effects on gameplay?

Same thing for plateaus: Aren't they just flat at a different altitude? I doubt that the transition from one level to the next will have any effect on gameplay, that couldn't be modelled by differences in altitude.

Of course artistically I agree that both would look cool. I just don't think they will have any great effect on gameplay.

As to plateus, yore right, except for the explicit example i made which does need to be modeled.

Some new things to add, mainly in referance to plate techtonis. Note that even though they may be underwater in our world and therefore mostly irrelavent, in other possible worlds this might not happen.
Base underlying landform:
Flat - Fairly even terrain, can be horizontal or sloping continuously. EX: Kansas plains
Hills - Uneven terrain with peaks and valleys, generally no rough terrain.
Foothills - Between Hills and Mountains. EX: Appalation mountains.
Moutainous - Jagged terrain with sharp peaks that are quite high. EX: Himalayas
Continental Shelf - Edge of the land part that is ontop of the plate in a geologically active world.
Valley - Opposite of hills, foothills, mountains. These are generally though not very sheer like mountains because of erosion and more along the line of the opposite of hills, though they can be on much larger extent. EX: Mediterainian (at one time this was also not underwater)
Trench - Formed along subduction zones these are extremely deep and sheer and long. Although physics on our world don't allow these to be above water as far as i know, it may differ for other worlds.
Ridge - Well not really a good term, but this is where there is a divergance in 2+ plates. Although generally below ground, they are found above ground (Africa). Ridges are lower than the rest of the and if it is on the landmass and higher if it is off the continetal shelf.
------
First off i'll generally skip over tempature since it isn't directly seen on the tiles, but it does have impact based on latitude, altitude and landshapes. I think its ness. to model all 3 in the ecology model somehow.

Water info: Water for map graphics will come, in general, in 4 types: Glacial/tundra, rivers, ocean/seas lakes/oasis and swamps/bogs.

glacial/tundra - the tempature here is cold enough for permafrost to exist. Even if its only the surface, this is what the player should see (this is that map graphics after all) We might haveto deel with the receeding and expansion do to seasons, but IMO we need ocean ice modeled because of restictions in movement (that's why the panama Canal was built...too long to go around S. America, and too dangerous most times to go around the north because of ice). Glacial retreat on land might cause glacial valleys if we want. Dunno bout the number of tiles, but atleast 1 for glaciers and 1 for tundra for several landtypes (tundra being just the snow and ice).

oceans/seas - in general there should be 2 types, shallow and deep. Shallow is for those on the contiental shelf and deep for those off. This doesn't include costal and transition tile layers. There could be a thrird for deeper parts on continetal shelfs, that aren't deep enough to be a deep ocean floor. It would be nice if these could be animated somewhat. Also oceans aren't ness salty.

rivers - these should be represented by blue lines of whatever width seems appropriate and whatever direction seems apporpriate for that tile. We shouldn't make too many variations and there can be things like meanders inside the tiles. Also rivers can create canyons, although whether a canyon or waterfall is created could be difficult to decide. Deltas can extend land and should be allowed to add land to the end and extend that land over long periods of time. A good example is the Tigris/Euphrates(?) valley. During the Mesopatamian era the delta was so far back the 2 rivers weren't joined. Now it is and a lot more land besides that has been clearly added. When new land tiles should be changed to reflect this i dunno, but this is not for this model to decide.

lakes/oasis - these are bodies of water that are small. These are less than 1 tile each by far (therefore the Great Lakes wouldn't be covered here, but on oceans/seas for modeling and art purposes. Lakes/oasis can vary in shape, but i dunno if its too good idea to do more than 1 shape per land tile layer except for flat lands and then only 2 maybe. Lakes can also exist as the beginning/ending or within rivers.

swamps/bogs - this also ties into vegetation a lot, but before vegetation can happen in this type of setting, water needs to be there. This is a place where the water table is above the ground level in many places, but trees, etc can grow easily. Also there are still many places that are still above the water table level, but the ground is saturated to an extent that it is always moist and usually difficult to walk thru do to the fact you can sink into the land itself.

That should be all. underground water isn't needing to be modeled neither rain (cloud modeling would be nice, but i doubt it will happen. Maybe underground water could be used for underground tiles. Right now this is focusing on surface area only.

vegtation: this area has tons of possibilities, more than any other. A few types have somewhat been mentioned in the other models. Tundra and swamp/bogs. There's also forests, deserts, savanas/plains, grasslands, jungles/tropical rainforest, brushlands, salt oceans (such as in austrailia), etc. And these can vary do to the tempature. FE forests in colder weather are usually pine forest (up to some extend both latitude and altitude), decidus in warm temperate (mixed in center) and other more equatorital areas there's other vegitation. Also some types of these can only exist on some type of land layers and water layers. A savana/plains cannot be on a mountain. Also there could be differnt denesites.

For the most part, underwater vegitation need not be worried about except maybe for reefs and kelp. Other areas are too small in area or don't add anything except frustration when creating.

Foothills:
If the only thing that distinguish foothills from hills and mountains are their altitude they don't need to be modelled, since altitude will be tracked separately in the mapsquare object. But if - as you say - it is worth modelling a radically different ressource component (one that couldn't be handled by adding specials or making small adjustments to the base of either mountains or hills) then foothills need to be modelled. IOW if foothills will be modelled separately by the movement, combat or ecology model then they have to have a unique map-graphic. I don't recall if the ecology model says anything about foothills. Do you think foothills will have unique effects on movement and combat, or will they just be a variation on the hills theme caused by different ground cover?

Continental shelf/shallow sea/deep sea:
Is the continental shelf then just a separator between shallow sea and deep sea (like a cliff or something) rather than the continental shelf itself.
I wondered: Isn't the underwater terrain on the continental shelf/shallow sea just variations on the aboveground terrain features, but covered with water. Mostly flatlands, but also some hills drowned by the ocean. Even mountaineous terrain is drowned by the sea: FE Norwegian fjords. In that case the transition between shallow sea and deep sea (what you call the continental shelf) is really a dividing line between landlike terrain and the oceanic floor. This would give the opportunity to model alterations in sea levels very easily. If we wanted it to be really beautyfull we could define graphics for partially ocean-covered hills, foothills or mountaiins. This would require that we define an average height of each terrain, and have a variable controlling sea level.
In that case I think the inclusion of underwater ridges and trenches will work fine for being the base terrains of the oceanic floor.
So what I am saying is that I think there are two different base terrains, with a number of different variations:
Landlike: Flat, hills, foothills, mountains
Oceanic floor: Flat, ridge, trench
The transition zone betwen the two is shown by a cliff
In theory all can be either above ocean level, partially above ocean level, or below sea level (each one requiring it's own graphic - the last one being the easiest ), although for earthlike scenarios the oceanic floor would be permanently submerged with the possible exceptions of volcanic activity in the middle of techtonic plates (hotspots like Hawaii). This could be modelled by a onesquare ridge (I know it's not accurate, but it is simpler and I think it will work). Another possibility would be to include another variation of oceanic terrain: Mountains.
Each sqaure should then have a base altitude, sqAlt, above some arbitrary zero level. Each terrain (flat hills etc) should have a base average height, teHeight, above the sqAlt:
If oceanic level < sqAlt then the square is above water level
If sqAlt < oceanic level< sqAlt + teHeight then the square is partially covered with water
If sqAlt + teHeight < oceanic level then the square is below sea level
In general I think the oceanic floor terrain types should not be allowed to have any vegetation, while submerged landlike squares could have kelp. I would disregard reefs for the moment.

Valleys:
I kind of figure valleys to be an integral part of the mountain/foothills terrain. If we want valleys approaching 100 km in width, isn't this just flatland (and likely a river) between two rows of mountains/foothills. I guess the Mediteranean and the Baltic could be seen as underwater valleys, but how do they differ from shallow sea/submerged land.

Water
Glacial/tundra: We have to model large bodies of ice off course. However, I'm not sure that tundra is a water type, rather a vegetation/groundcover type. For the moment I would forget about modelling the retreat or growth of ice sheats on land. At sea, however, changing sizes of the ice cap with the season would be nice, and easily modelled. I don't think that the distinction between salt amd freshwater oceans needs to be covered graphically. It might however be necessary to model it in the ecology model.

Sea/Ocean: If done as above the distinction would no longer be between shallow sea and deep sea, but between ocean above ocean floor, and ocean above submerged lanlike terrain. Basically they could be shown as two different hues of blue.

Rivers: IMO opinion forget about canyons and waterfalls, concentrate on deltas, which have had large influence on civ in some places (notably Egypt, Mesopotamia).
BTW I know it will not look pretty on a square-based map, but have you ever discussed having rivers and such along the square-edges, instead of in the square. This is rather attractive on hexbased maps, and might give some additional opportunities regarding bridges, rivercrossings etc. I will make an inquiry in the Mapsquare OO thread re hexmaps.

Lakes/Oasis: I agree with lakes, but again i think oasis is more like a vegetation type, only existing in desert areas where there is a local access to surface water or relatively easy access to underground water sources.

Swamp/bogs: Again more like groundcover

quote: Foothills: If the only thing that distinguish foothills from hills and mountains are their altitude they don't need to be modelled, since altitude will be tracked separately in the mapsquare object. But if - as you say - it is worth modelling a radically different ressource component (one that couldn't be handled by adding specials or making small adjustments to the base of either mountains or hills) then foothills need to be modelled. IOW if foothills will be modelled separately by the movement, combat or ecology model then they have to have a unique map-graphic. I don't recall if the ecology model says anything about foothills. Do you think foothills will have unique effects on movement and combat, or will they just be a variation on the hills theme caused by different ground cover?

quote: Continental shelf/shallow sea/deep sea: Is the continental shelf then just a separator between shallow sea and deep sea (like a cliff or something) rather than the continental shelf itself. I wondered: Isn't the underwater terrain on the continental shelf/shallow sea just variations on the aboveground terrain features, but covered with water. Mostly flatlands, but also some hills drowned by the ocean. Even mountaineous terrain is drowned by the sea: FE Norwegian fjords. In that case the transition between shallow sea and deep sea (what you call the continental shelf) is really a dividing line between landlike terrain and the oceanic floor. This would give the opportunity to model alterations in sea levels very easily. If we wanted it to be really beautyfull we could define graphics for partially ocean-covered hills, foothills or mountaiins. This would require that we define an average height of each terrain, and have a variable controlling sea level. In that case I think the inclusion of underwater ridges and trenches will work fine for being the base terrains of the oceanic floor. So what I am saying is that I think there are two different base terrains, with a number of different variations: Landlike: Flat, hills, foothills, mountains Oceanic floor: Flat, ridge, trench The transition zone betwen the two is shown by a cliff In theory all can be either above ocean level, partially above ocean level, or below sea level (each one requiring it's own graphic - the last one being the easiest ), although for earthlike scenarios the oceanic floor would be permanently submerged with the possible exceptions of volcanic activity in the middle of techtonic plates (hotspots like Hawaii). This could be modelled by a onesquare ridge (I know it's not accurate, but it is simpler and I think it will work). Another possibility would be to include another variation of oceanic terrain: Mountains. Each sqaure should then have a base altitude, sqAlt, above some arbitrary zero level. Each terrain (flat hills etc) should have a base average height, teHeight, above the sqAlt: If oceanic level < sqAlt then the square is above water level If sqAlt < oceanic level< sqAlt + teHeight then the square is partially covered with water If sqAlt + teHeight < oceanic level then the square is below sea level In general I think the oceanic floor terrain types should not be allowed to have any vegetation, while submerged landlike squares could have kelp. I would disregard reefs for the moment.

quote: Valleys: I kind of figure valleys to be an integral part of the mountain/foothills terrain. If we want valleys approaching 100 km in width, isn't this just flatland (and likely a river) between two rows of mountains/foothills. I guess the Mediteranean and the Baltic could be seen as underwater valleys, but how do they differ from shallow sea/submerged land.

quote: Water Glacial/tundra: We have to model large bodies of ice off course. However, I'm not sure that tundra is a water type, rather a vegetation/groundcover type. For the moment I would forget about modelling the retreat or growth of ice sheats on land. At sea, however, changing sizes of the ice cap with the season would be nice, and easily modelled. I don't think that the distinction between salt amd freshwater oceans needs to be covered graphically. It might however be necessary to model it in the ecology model.

quote: Sea/Ocean: If done as above the distinction would no longer be between shallow sea and deep sea, but between ocean above ocean floor, and ocean above submerged lanlike terrain. Basically they could be shown as two different hues of blue.

quote: Rivers: IMO opinion forget about canyons and waterfalls, concentrate on deltas, which have had large influence on civ in some places (notably Egypt, Mesopotamia). BTW I know it will not look pretty on a square-based map, but have you ever discussed having rivers and such along the square-edges, instead of in the square. This is rather attractive on hexbased maps, and might give some additional opportunities regarding bridges, rivercrossings etc. I will make an inquiry in the Mapsquare OO thread re hexmaps.

quote: Lakes/Oasis: I agree with lakes, but again i think oasis is more like a vegetation type, only existing in desert areas where there is a local access to surface water or relatively easy access to underground water sources.

quote: Swamp/bogs: Again more like groundcover

quote: Come to think of it these water-objects (including ocean) should be handled just like other groundcover/vegetation objects, with the possible exception of rivers, since they have connections to neighbouring squares and are more akin to infrastructure. There are some important OOA implications of this, so if this is the way to do it we would have to integrate it into the Mapsquare OOD, which should be very easy.

Anyway can u give some suggestions on the other 2.

The code right now has 'terrain' as just the representation of how 'broken' the ground is. So there's going to have to be 'grades', like in SMAC -- which has 'flat', 'rolling' and 'rocky'. The square will have 'elevation' values.

'Water' will be a 'resource' on the terrain.

So a 'valley' within one square will be easily done by something like 'terrain = hills'. A multi-square 'valley' like the Mediteranean would be done via 'elevation' values in mapsquares.

Then a valley could have 'water' objects in the squares. The 'water' objects would have depth, width/% of square covered, that kind of info.

This also means that an 'ocean' square is easily modelled. Just define the world's terrain first, then set a 'sea level'. Then fill in 'water' objects on all mapsquares that are both lower than sea level and connected to other filled squares.

Can you explain grades better. Also i don't think there should be too much emphasis here because already with what i have proposed there will be tons of tiles or tile layers maybe better sounding?

'Grades' is just a name I came up with. The terrain object describes the different types of 'Earth's Crust' that might be in each mapsquare.

I think ya'll game designers should define them. I'm just telling you we'll need them.

I would think it could be any number of different things, tho -- volcanic, flat, fault, rolling, rocky/mountainous, etc. Maybe none of those. It could be expanded to include 'lake bed', 'river valley', 'mountain pass', 'plateau', or anything, later. Ya'll decide.

The idea is that a mapsquare will be a combination of 'crust' (the terrain object) and water, if present.

I find it very hard to make any judgement based on only 3 tile types. The grassland (?) and coastline look pretty good to me, and the water is I think too boring.

Others:

I'd prefer regular hexagons.

Why not use these temporarily for Clash?

We would, of course, get the owner's permission, then resize them to whatever size we need. This would give us some graphics to work with until we have artists on board to do all the different graphics for the final game. In all I don't think it would be too much work to get the 2 different sizes for all tiles we need, since I don't think we'll need all those unit graphics. If you like the idea, I'll need a list of what units, terrain, and city icons we'll need (with different cultures, etc.).

The status for now is that we are still using GIF 89 a tiles. That will probably have to end at some point, since GIF licenses are being required for commercial projects, and they cost a non-trivial amount of money. But for the moment I think we can do a lot of good exploration using the GIF format, and hopefully move significantly forward and lock in specifications for the graphics.

Why can't we change the tile size we are using? It would be trivial, and it would allow us to use any existing Civ2 graphic that the owner gives us permission to use. Personally, I like the civ2 tile size better than the larger ones we are currently using.

On changing the tile size... Well I think the bigger ones look better, simply because you get more pixels to improve the appearance. But you are right that changing our default tile size would allow for immediate use an awful lot of already-done civ 2 stuff providing we can get permission. Hopefully we can get some of the artists to weigh in on whether the graininess existing civ ones would be a problem from an aesthetic perspective...