1. We have added the ability to collapse/expand forum categories and widgets on forum home.
    Dismiss Notice
  2. Photobucket has changed its policy concerning hotlinking images and now requires an account with a $399.00 annual fee to allow hotlink. More information is available at: this link.
    Dismiss Notice
  3. All Civ avatars are brought back and available for selection in the Avatar Gallery! There are 945 avatars total.
    Dismiss Notice
  4. To make the site more secure, we have installed SSL certificates and enabled HTTPS for both the main site and forums.
    Dismiss Notice
  5. Civ6 is released! Order now! (Amazon US | Amazon UK | Amazon CA | Amazon DE | Amazon FR)
    Dismiss Notice
  6. Dismiss Notice
  7. Forum account upgrades are available for ad-free browsing.
    Dismiss Notice

Concept for a C2C Climate System

Discussion in 'Civ4 - Caveman 2 Cosmos' started by DRJ, Apr 15, 2012.

  1. DRJ

    DRJ Hedonist

    Joined:
    Dec 1, 2005
    Messages:
    1,514
    Location:
    Hamburg, Germany
    Concept for a use of the property system towards

    rainfall & river patterns interacting with wind patterns interacting with ocean currents interacting with sealevel (bound water in ice) patterns that then interact with the rainfall patterns again.


    1.

    A modified "storm" (rainclouds) could be happening on land tiles.

    It would increase the density of that tile and the tiles it surrounds.

    A high density level favors forming of rivers which are not static anymore.

    Density, like crime, could drip towards the sea tiles via an invisble height variable, so the rivers seem to be natural, though changing upon the landscape and sealevel.

    The more density a tile gets, the more it changes status over time.

    Lowest density is salt flats/dunes in warmer regions and barren in cooler climate.

    Highest density is muddy & jungle in warmer climate and tundra & swamp in cooler climate.

    Rainfall patterns will be able to change because the wind pattern changed because the ocean currents changed because the sealevel changed because some vulcanoes erupted (lowering gloabal temperature - a new variable) and because the axis of earth changes slightly back and forth over time (another to be implemented static variable, like latitude).

    2.

    wind patterns
    direct the rainfall.
    (wind, like crime, drips over land tiles towards hills and mountains, where it accumulates and that favors a triggering of 'land storm' aka rainfall in front of ridges)
    and the speed of ships.

    3.

    ocean currents direct the wind
    as well as the some new static variables like and local seasonal effects because of the seasonal rotation of earth axis (jetstream/trade wind).

    ocean currents tend to lose their rotation speed the closer they are to land ->
    in middle of ocean the broadest currents, which are triggered to organize in maelstroms, appear.


    4.

    a new sealevel variable, dependent of vulcanic activity, longtime rotation of the axis of earth and perhaps other, yet to be introduced variables,
    re-directs ocean currents.

    If sealevel variable loweres, the new coastal lines will allow currents to change and bring density to other regions which will
    produce ice age effects (more water bound in ice=less density more and cooler wasteland) and warmth period effects (if sealevel rises again).


    The one problem I see here is that the new land would be lower than the old coast land which would be higher in relation to the new land. So a invisble variable like Height would help to cope with that.
    The lower the sealevel gets, the more flat land has to become hill, in relation to the new coastline.

    So for example you have a grassland coast that will stay flat tile once the sealevel lowered and a new terrain forms on the old coastal waters -- ice? in artic, salt marsh (same graphics like swamp maybe) in temperate and mangroves in tropical. Once the sealevel falls another tile short, the old salt flats would become grassland and the old grassland would become hills. Inland, the flats would stay flat like tableland. To avoid a "ring of hills" at the old coasts, some of the grasslands would not become hill, it may be a 50% chance. Rivers could also drip towards ocean by using the height variable, floating through the old coastal grasslands that didn't become hills into the sea.


    Another variable that could and maybe even has to be introduced into this climate simulation would be the local saltlevel in the water as it drives the currents. Any ideas about that?


    One thing that bothers me a bit is the spread of organic resources. They, somehow had to become dynamic resources as well, adapting to shifting landscapes.

    Also the improvements of terrain. What happens to a farm when the grassland becomes swamp?

    One thing would be that the more the game progresses the less would change per turn as the timeframe of the turn 1 year compared to 1000 would be like a brake for the earlier more frequent changes. If the start would be 1 Mio BC, by the time you get to 6000 BC you can get some constant agriculture going (maybe change the turn-costs for making improvements by relating it to the numbers of landscapeshifting effects active/turn) but of course it can be washed away by floods like in sumeria, or even like the black sea had its great flood. Early cities near the water? not so good idea maybe, although you might be spared. But building them in safe terrain with low density might not help them to grow fast... choices over choices^^ (being able to be killed by nature as a new optional... :rolleyes: feature hehe)

    Please post ideas and suggestions so this can be discussed for planning of steps towards (all optional?) realization
     
  2. Thunderbrd

    Thunderbrd C2C War Dog

    Joined:
    Jan 2, 2010
    Messages:
    19,838
    Gender:
    Male
    Location:
    Las Vegas
    I really like most of this. I had an idea before for how we could free up animals to be a bit more dynamic on the map. Very interesting concepts you're laying out all in all. My only concern would be the massive amount of data processing it would add... I turn storms off as they are already to help with the game speed.
     
  3. Koshling

    Koshling Vorlon

    Joined:
    Apr 11, 2011
    Messages:
    9,056
    Anything that chnages domain ype is VERY expensive to process (water <-> land). Anything that changes trade routes (river movemn) is pretty expensive and should not happen frequently.
     
  4. DRJ

    DRJ Hedonist

    Joined:
    Dec 1, 2005
    Messages:
    1,514
    Location:
    Hamburg, Germany
    The early times when the terrain changes frequently there won't be many trade routes but mostly nomad tribes.

    The trade between them could be organized by hand to hand (trade possibility established with unit on unit), so the processing for normal trade routes would only be established once the game has become more and more static, which will make no need for reloading of "normal traderoutes" until one gets barter/trade tech at which time the process will be called the first time.

    The approach here is to make what we can out of the limited RAM resources at every era.
    As long as world is relatively unpopulated the RAM resources for later buildings, trade routes etc can be used to process the new variables I listed above. So if we can have it under full steam at all time without having it boil over thats the perfect situation I think. At least for Super Eternity, (right, Koshling? :D)

    We could also fake the water becoming land process somehow by having a more or less predetermined min/max coastline upon map start, which would be able to look like water and have sea critters swimming on it but which is "land" in code, just shifting graphics.
    Lot of ways to improvise there.
     
  5. Acularius

    Acularius Chieftain

    Joined:
    Jan 22, 2012
    Messages:
    724
    Location:
    Ontario, Canada
    How static could these changes be made, in terms of scenario creation and pre made maps?

    I do enjoy such things as volcanoes already, but I am just curious on how many changes to 'Earth' it might create.

    I must say that I find it intriguing. Although map creation is going to require MUCH more data mining. XD
     
  6. Azurian

    Azurian The Azurian

    Joined:
    Apr 10, 2012
    Messages:
    2,092
    Location:
    Florida, USA
    I found something of interest.

    Accurate Earth Map
    http://forums.civfanatics.com/showthread.php?t=233586

     
  7. DRJ

    DRJ Hedonist

    Joined:
    Dec 1, 2005
    Messages:
    1,514
    Location:
    Hamburg, Germany
    Perhaps some changes like sealevel could be optional predetermined, having an illusion of a living map but it is bound on certain parameters determined at map start, making the whole thing more easy to process.

    This approach (which I dislike a bit for random maps as I favour unpredictability there) may work for GEM where you could use the code of worldbuilder about which tiles are seen by player and which not to make another worldbuilder option that lets you define which tiles are changed with different sealevels and which not.

    So you could for example predetermine the North Sea would become land if the sealevel drops.



    @MrAzure Yeah I posted that a while ago but everybody had their hands on something else and I was not able to implement anything. Btw. would you like to help with this? I definitely need more help to picture this and organize possible code etc... Maybe I can't do anything besides this concept at all, because I have no idea how the property system works; once I understand that I maybe can lay some basic roots but I think that will depend on the real professionals^^
     
  8. Hydromancerx

    Hydromancerx C2C Modder

    Joined:
    Feb 27, 2008
    Messages:
    16,244
    Location:
    California, USA
    One problem is the "Storms" that appear in the ocean are actually Terrain Features same as a Forest or Jungle. The problem is adding any weather to a layer of the map without replacing terrain features like the jungle or forest. It could possibly be done using units instead since they can appear on Terrain Features. The problems is not having units "attack" the weather.
     
  9. DRJ

    DRJ Hedonist

    Joined:
    Dec 1, 2005
    Messages:
    1,514
    Location:
    Hamburg, Germany
    lol I know a unit attacking weather - it's called HAARP.
    Or imagine the russian fighter jets that kill the clouds before parades are held.

    But regarding that solution I like it!
    If the rain(man) would just have to be a visble (storm graphic) spy type unit who carries no flag with it -
    (can that stupid flag icon not be disabled or altered in any way?- I need to know for rescaling it, there must be an icon code somewhere, mustn't it?)
    and the opportunity fire it lays on the landscape produces the density (much like fallout lol)

    In fact maybe the combination of simulating climate processes with modified invisible units is quite charming and can interlock nicely with glacier tongues etc. Some of you may have seen the glacier tongues I artificially created in here:
    Spoiler :
    It's only recreatable for screenshot purposes atm. ice"units" may change that. (these "icemen" will then smelter in the sun at some time^^)
     
  10. AIAndy

    AIAndy Chieftain

    Joined:
    Jun 8, 2011
    Messages:
    3,352
    Units don't really work for that kind of thing (due to the way they are displayed).
    But it is possible to have more than one feature on a plot. That just requires some code changes that I want to do anyway to have properties influence plots (similar to the auto buildings/promotions that we have now).
     
  11. Hydromancerx

    Hydromancerx C2C Modder

    Joined:
    Feb 27, 2008
    Messages:
    16,244
    Location:
    California, USA
    That's great news AIAndy! :D It would be great to have a "Weather" layer to the map. :D
     
  12. AIAndy

    AIAndy Chieftain

    Joined:
    Jun 8, 2011
    Messages:
    3,352
    Mind though that while I generally think dynamic climate and weather systems would be very interesting, there are quite some problems associated with it:
    • Improvements are restricted to certain terrain types and it is not quite clear how that would interact with changing those types semi-often
    • Climate simulations are expensive. The generic property system propagators are only solved once per turn but proper climate simulations would require solving on sub time steps to get proper results.
    • Having the ability to influence climate by certain actions can be used for war but it is very difficult to teach this to the AI
    • The map scripts need changing to get the information that is needed for proper climate simulations (several of them do a kind of climate simulation of their own)
     
  13. DRJ

    DRJ Hedonist

    Joined:
    Dec 1, 2005
    Messages:
    1,514
    Location:
    Hamburg, Germany
    Yes, improvements would be destroyed periodically. But consider this: in nomad time there are gatherers that move on with the nomad camp once it moves on. No gatherer improvements like now but just pure gathering (aside of trails that may be washed away by flooding perhaps). At about 8000-to 5000 BC the first farms will be in danger. Rightfully so. They haven't learned canal systems and stuff yet.
    Turn cost for building improvements in that time could be associated to how much weather changes. So it would be cheaper to build an 'early' farm than a 'later' farm (a later farm will give more boni anyway thx to the techs like plough!).

    For example:
    around 6000 BC average of 20 weather (property) events in the world @ that turn? -4 turns for building a farm.
    around 0AD 5 weather events /turn? -1turns for farm.

    I made a sketch about how I see the processor load could be allotted -- the sketch is not very accurate and more like a wish, I guess but it may help seeking potentials.
    I encourage to make own sketches so the idea of a constant high level of processor load throughout the game can be planned better.

    Yes but as the climate effects will become seldom (until industrial age and a potential new climate change for pollution but that's not certain) there won't be much reason to have great AI coding for this. I mean yes, regarding weather warfare for example the US have plans to have some artificial storms hammer the venzualians coast for two weeks before marching in so the resistance can't be effective in all the mud but that's another thing for much later maybe.

    Can't the maps climate system be used for general manifestation of a map and the new weather system pick up from there on from turn1? I don't see why that would contradict. After all the different simulations would complement each other, leading to interesting effects.
     

    Attached Files:

  14. Koshling

    Koshling Vorlon

    Joined:
    Apr 11, 2011
    Messages:
    9,056
    Have you looked at CPU load while running C2C in Performance Monitor? It slams one core totally (100% usage throughout the AI turn) already. Without making things multi-threaded (hard in general) anything added is longer turns, since there is precisely zero slack.
     
  15. DRJ

    DRJ Hedonist

    Joined:
    Dec 1, 2005
    Messages:
    1,514
    Location:
    Hamburg, Germany
    Ok, I haven't done that yet (how can I run it with a performance monitor?).

    Well I know weather would come at a certain cost. All I wanted to picture was that by making some processed factors like traderoutes start using process time only in later ages instead of from the very beginning there would be a little room to fill for weather system, or at least the increase of turntimes wouldn't be so dramatically.

    Basically, in later game there should be no or just a 10 % increase with weather. For early game, well I wouldn't bother a few more secs per turn as thx to your optimizations, Koshling, my treshold of frustration about turn times has been reduced to zero atm, so I am pretty open for longer turntimes (as in combination with autoendturns this might give the feeling of a 'living world' in a turn-based game, something civilization has never had before and which was my idea in the first place)
     
  16. Koshling

    Koshling Vorlon

    Joined:
    Apr 11, 2011
    Messages:
    9,056
    I just meant start task manager, and select the 'performance tab' to see CPU usage by core. My point however, was that C2C already fully itlises the one thread it has with no slack. That's not to say we can't have background threads processing on-going things like climate models, it's just not easy because the whole foundation is built for a single-threaded environment.
     
  17. DRJ

    DRJ Hedonist

    Joined:
    Dec 1, 2005
    Messages:
    1,514
    Location:
    Hamburg, Germany
    Yeah background threads, that sounds nice. How are they established?
    I was thinking if there can't be a power transformator module, that converts from single threaded into multithreaded like co-current flow becoming single-phase alternating current...

    btw once you're asked by someone what youre doing as a profession you can say
    "I'm a backdoorground thread man" :cool: ^^


    EDIT: so maybe the Y axis of my sketch from above should not have % of processor load
    but turn loading time instead -- what the sketch could provide would be an example for a strategic oversight of different evolving factors during a game that allows to compose relatively average loading turns throughout the game.
     
  18. Koshling

    Koshling Vorlon

    Joined:
    Apr 11, 2011
    Messages:
    9,056
    How shall I put this....
    Spoiler :
    lol
     
  19. DRJ

    DRJ Hedonist

    Joined:
    Dec 1, 2005
    Messages:
    1,514
    Location:
    Hamburg, Germany
    hey, come on! now you're mean :cry:

    How on earth shall I know how all of this stuff really works? The knowledge of this is slowly soaking through the soil of my god-mode-concepting-ignorance.

    Because as we say in Germany: "Hilf dir selbst, dann hilft dir Gott"
    -- help yourself, then god will help you^^
     
  20. rightfuture

    rightfuture Chieftain

    Joined:
    Mar 11, 2008
    Messages:
    1,649
    me too :)

    I've talked about kevinman4400's accurate earth map, and world history mod projects a bunch in the C2C forums. I think basic modeling of weather, since it affects history so much, would be a nice major focus for future modding. I like the idea of using basic fractal math (like the terrain scripts) to simulate simple moving storm fronts. Gradual climate shifts also radically altered human cultures. Check out the fall of the Mayan civilization for one.

    Also note
    C2C -Weather discussion
    http://forums.civfanatics.com/showthread.php?t=439763
    I have tried to start ongoing discussion on weather before.

    I mention the Ice Age script for receding glaciers from FFH -Fall from Heaven mod every so often. it could be adapted without terrible effort I think. Sea levels in receding Ice Age times changed a good bit also, even in the last 14,000 years; Has anyone noticed how many underwater prehistoric city discoveries have been made in the last 10 years(from Egypt, Japan, Black Sea, North Sea, Great Lakes, etc.) ? Google it and you'd be suprised.

    Weather and disease are among the biggest things that have historically shaped the development of human civilization. It think it is well worth the time for the C2C modders to consider focusing on them after after recent plans pan out.
    Can't wait to see a little development on those in C2C . Modders please consider making it part of your near-future plans. :)

    Like DRJ I think it is worth some simple modeling and performance cost.
     

Share This Page