Dynamic Terrain for RFC

[Panopticon]

http://forums.civfanatics.com/downloads.php?do=file&id=13466

One last modcomp before I go on a trip to London. This modcomp changes the terrain in certain parts of Asia and Africa as the mod enters the colonial era. Jungle in city sites like Manila, Colombo, Lagos and Bandar Seri Begawan will turn into forest in 1600 AD. The Indus Valley will change from unadorned desert to flood plain, so that Karachi can grow. There are also extra dynamic resources added to the map, in order to make the cities viable.

At the moment, many of the world's current significant metropolitan areas are uninhabitable on the RFC map due to terrain features. This is good for reducing the growth of early civs, particularly India, but it prevents the growth of cities like Karachi and Manila which became important more recently in history. I don't know if the AI will settle the new areas to any extent. Consider the mod in version 0.1 at the moment; any suggestions for new city sites are welcome.

Thanks to NerfCothons and the WASP mod for the inspiration and most of the detail positioning: http://forums.civfanatics.com/showthread.php?t=328203

The zip file simply contains a modified version of the Resources.py file located in Assets\Python. Replace the existing Resources.py file with this one.

 

[NBAfan]

This is very nice! You now have a download.

 

[killerkebab]

I really, really, really like this. Is it also possible to "entice" the AI to settle cities such as Bogotá, La Paz, Manaus, Panama? I notice in central Africa as well, there are places that look interesting (I'm thinking there's a lone grassland plot in the middle of the jungle with bananas and some metals nearby), whose real life city I don't know the name of - but no AI ever settles it.

 

[Panopticon]

I really, really, really like this. Is it also possible to "entice" the AI to settle cities such as Bogotá, La Paz, Manaus, Panama? I notice in central Africa as well, there are places that look interesting (I'm thinking there's a lone grassland plot in the middle of the jungle with bananas and some metals nearby), whose real life city I don't know the name of - but no AI ever settles it.

Thanks. I'm sure it's possible to do that by changing the settler maps, which are contained in the DLL, I think. So it would require either a recompiled DLL file or some kind of map-altering Python function, and I don't know if the second proposal is possible.

 

[evran]

Downloaded it

 

[killerkebab]

Bombay grows, but not nearly as much as it should imo - if I recall correctly, real life Mumbai is one of the largest cities in the world today by population, however because it is surrounded by plains rather than grassland it doesn't really grow in the modern age. I'm not suggesting we should add grassland, but maybe more food resources. What date did you say the resources should show up? I am at 1950AD at the moment, maybe you set it so Bombay gets a growth boost later

The Indus Valley changed to flood plains as you planned, however no AI seems interested in Karachi

Lastly, though Colombo is now forest rather than jungle, it doesn't seem to appeal to the AI.

Lagos gets settled by the AI in almost 100% of my games - usually by the Portuguese.

In some of my games the Khmer decide to settle Brunei, but it is a much rarer occurance than it is in the WASP mod (I assume this is because in the WASP mod the tiles are immediately settle-able). The Japanese rarely settle Manila. They usually go for Raoagu and Davao, although in a couple of my games they did settle Manila and start chopping forests when they reached biology. Any more spots I should be checking?

 

[Panopticon]

I have noticed:
- Southern Nigeria is settled with > 80% probability. Usually by the Portuguese.
- The Karachi-Gujarat-Mumbai area is used by either Turkey or Independents (Mumbaiin 600 AD start) with 80% probability, but the Indus Valley itself isn't settled often.
- Khmer settles Brunei with 50% probability.
- Japan settles Manila with 20% probability. It does so if it has a presence on the Philippines already.
- No-one cares about Sri Lanka.

killerkebab, I think Bombay gets a bonus fish resource in 1600 AD and that's it.

 

[killerkebab]

I believe in the late game there simply isn't enough food to boost the big population centres such as Colombo and Mumbai. If, say 100 turns from the game end, more food appeared next to today's big population centres - would the result damage the game mechanics?

 

[Panopticon]

No it wouldn't.

 

[NerfCothons]

Just realized you had done this Panopticon, and I'm a fan. Bump!

Lagos is definitely the clear winner in this mod. Mumbai would be as well if only we could convince India to settle it's cities correctly.

Another mechanic I've been considering recently would be desertification. We would have more freedom to make Egypt more appealing earlier in the game if those flood plains gave way to desert over the course of the game.

 

[Baldyr]

Another mechanic I've been considering recently would be desertification. We would have more freedom to make Egypt more appealing earlier in the game if those flood plains gave way to desert over the course of the game.

Yeah, and Northern Africa shouldn't be as arid in ancient times.

Some code that takes an area of the map and degrades all tiles grassland -> plains -> desert would probably do the trick. Then it would just be a matter of defining those areas and have the code trigger at key dates.

A more refined code would only degrade every other tile (determined randomly), giving less predictable but still foreseeable effects over time. So if you "degrade" the same area three times you would pretty much have all desert, with some plains thrown in and possibly a few green grassland oasis also. But nearing modern time there could be a fourth and a fifth cycle...

I actually think I'll add this functionality to the scenario maker application I'm doing right now.