Desert tile should be floodplain

[Vladesch]

I thought any desert tile next to a river should be floodplain.
The selected worker in this picture is standing on desert, but shouldnt it be a floodplain?

Here is the savegame...
desert.CivBeyondSwordSave

 

[playshogi]

This was always in the game. Was this square far north or far south? It's only a floodplain in tropical or temperate regions, not polar ones.

 

[Vladesch]

Ahh wasnt aware of that restriction.

I know I was well north of the equator so I guess that explains it.
Moderator can mark this thread as *not a bug* please.

 

[ainwood]

This was always in the game. Was this square far north or far south? It's only a floodplain in tropical or temperate regions, not polar ones.

I'm not convinced by that - and if you look at that save, its not particularly far north, and there are floodplains that are further north.

Was this a river added in worldbuilder? Or was it just randomly generated like that?

 

[Bindamel]

Can't open the save right now, but does the river pass by a capital? Sometimes the game adds a river to a starting location after the world is built, and does not call the function to add flood plains, so you're left with desert tiles next to rivers. It's rare, but has existed since Vanilla.

 

[Roland Johansen]

I've read that this is being caused by the following bug-like behaviour.

The map generator creates a map and all the desert tiles along rivers are floodplains. Then the civilizations are placed on the map. The area around the settler of all of the civilizations is improved so that every civilization starts in good terrain. Sometimes this includes adding rivers to the map. If those rivers happen to pass through desert tiles, then those tiles are not upgraded to floodplain tiles.

Is the river in question somewhere on its route close (within 5-6 tiles) to a capital?

By the way, I've read this. I don't know if it's true. I do know that I've seen deserts along rivers in earlier versions of civ4, pre-BTS. In all cases they were next to rivers that were close to the capital or were caused by the destruction of a city that was founded on top of a floodplain tile. (Floodplains are overlays and overlays are removed by settling a city, just like forests and jungles.)

The question is, should the algorithm that improves the land around capitals and thus sometimes places rivers also upgrade deserts to floodplains? It would then upgrade the terrain a lot further than by just adding the river. From a game consistency point of view, I would say yes (deserts next to rivers are floodplains everywhere). From a balancing starting positions point of view, I would say, I'm not sure.

I would personally prefer if the starting position algorithm would add the floodplain overlay to deserts next to rivers. I would also prefer if the floodplain overlay was added again when a floodplain city was destroyed. I like consistency in a game.

edit: Bindamel was faster.

 

[IronCrown]

Actually, there are deserts next to rivers in the real world. It depends on very rare geographical features, but it's possible.

 

[Andraeianus I]

Is there a way to change the behaviour of the map generator in order to have always floodplains instead of desert next to rivers?

 

[Olson]

I've read that this is being caused by the following bug-like behaviour.

The map generator creates a map and all the desert tiles along rivers are floodplains. Then the civilizations are placed on the map. The area around the settler of all of the civilizations is improved so that every civilization starts in good terrain. Sometimes this includes adding rivers to the map. If those rivers happen to pass through desert tiles, then those tiles are not upgraded to floodplain tiles.

Is the river in question somewhere on its route close (within 5-6 tiles) to a capital?

By the way, I've read this. I don't know if it's true. I do know that I've seen deserts along rivers in earlier versions of civ4, pre-BTS. In all cases they were next to rivers that were close to the capital or were caused by the destruction of a city that was founded on top of a floodplain tile. (Floodplains are overlays and overlays are removed by settling a city, just like forests and jungles.)

The question is, should the algorithm that improves the land around capitals and thus sometimes places rivers also upgrade deserts to floodplains? It would then upgrade the terrain a lot further than by just adding the river. From a game consistency point of view, I would say yes (deserts next to rivers are floodplains everywhere). From a balancing starting positions point of view, I would say, I'm not sure.

I would personally prefer if the starting position algorithm would add the floodplain overlay to deserts next to rivers. I would also prefer if the floodplain overlay was added again when a floodplain city was destroyed. I like consistency in a game.

edit: Bindamel was faster.

[Note: Before the anti-necro crowd spams this post, I got a (d20, 15+)x2 against anti-necro attacks....I'm supposed to be safe.]

In addition to the quoted text, but also explained by it, it's possible for grassland and plains tiles next to the aforementioned river type to be unfarmable, despite being tagged as 'fresh water', until it's chain-linked (haha!) after Civil Service. I discovered this in a game last night and a thread search found the above. Just bringing it up for interesting reading and as a FYI. Chill. No Reply Requested.

 

[neilmeister]

Anti-necro crowd? ... I'm sorry I have no idea what you are talking about

 

[Olson]

Anti-necro crowd? ... I'm sorry I have no idea what you are talking about

I'm referring to those Beavis & Butthead types who don't know the difference between a *dead thread* (one which is time sensitive and therefore no longer relevent) and a *dormant thread* (one which is still relevent but hasn't seen recent activity).

 

[plastiqe]

Olson your anti-necro saving throw negates all damage. : P

I have seen instances of river tundra forest that doesn't get with a Lumbermill. I assume either because the forest grew there or it's the same rivermaker issue.