Future Direction of the Mod

[NKN]

maybe even a new, standalone game that could be started from scratch

 

[Hippo8085]

Out of curiosity, how set in stone is the way AI settling patterns works? The more I think about it, the more I am favoring sort of a small regions (provinces, if you will, but only 4 -12 tiles each) approach to how the AI settles the map, with more generic maps. The same could probably be approximated with the current system, but I don't think it would work quite as nicely.

More on this if you think the system could change...

 

[Leoreth]

If we want to change the fundamental ways the original Civ4 engine works, that should either be a separate project or maybe even a new, standalone game that could be started from scratch (of course, the latter is a pipe dream that Leoreth probably doesn't have the time or motivation to pursue).

*sets up a Kickstarter*

Out of curiosity, how set in stone is the way AI settling patterns works? The more I think about it, the more I am favoring sort of a small regions (provinces, if you will, but only 4 -12 tiles each) approach to how the AI settles the map, with more generic maps. The same could probably be approximated with the current system, but I don't think it would work quite as nicely.

More on this if you think the system could change...

Not sure what you mean by that. Everything about the mechanic can be changed, but that doesn't mean that everything is possible. AI is a complex interdependent system.

 

[ales_]

BTW, the last version without a civ was 1.81 but the 1.xx versions were usually bugfixes. If you don't count them, the last version is 1.5, before the addition of any civ!

 

[Leoreth]

Yeah, I didn't count them. Interesting though!

 

[JiimBV]

I think Hippo is asking (correct me if I'm wrong) if it would be possible for DoC to take on a SoI style province system and incorporate AI settling patterns into provinces, rather than individual tiles. Because of this I think AI settling could become more intuitive and perhaps more spread out. Though I don't know very much about the AI I admit.

 

[Leoreth]

The SoI AI uses the same settler maps that DoC and every other RFC mod uses.

 

[need my speed]

But you don't have to end there, you can make civics depend on others or have two civics being incompatible.

I view such a system as your civilization's skill tree more than the traditional set of specialization options. There could be a "republican tree" and a "monarchist tree" for instance with different civics associated with it - adopting new civics within a tree could be free while switching would produce anarchy and so on. There are a lot of directions you can take this into.

Indeed, and I have toyed with this, but it gives you the same result I guess. You have a 'democracy' tree, a 'monarchy' tree, an 'authoritarianism' tree, and an 'oligarchy' tree, for example, and within are the relevant options ('constitutional', 'absolute', et cetera for monarchy). But that's only talking about the government.

It could work. You'd introduce some kind of tab system, so the basic civic screen of IV is merely one tab, called 'government', on which you see the above options. Go to the tab 'warfare', and you see war related options. However, I can't think of any options that can be so detailed as the above 'government'.

Something like V's system also doesn't really fit, because that is less about forms of governments and more about concepts attached to one core concept / ideology. And the progression doesn't make a lot of sense.

Something like Europa Universalis would become insanely complex, even if you remove 75% of the options present n that game (because you're covering a lot more of history, so you'd have a lot more 'civic' options). You'd need to take a hundred steps to go from 'absolute monarchy' to 'direct democracy' (as that is how Europa Universalis works, I believe; you switch from 'absolute monarchy' to 'constitutional monarchy with parliamentarian democracy' to 'elected head of state with parliamentarian democracy' to 'elected head of state with direct democracy', for example). And, it just doesn't appeal to me; 'elected head of state' and 'constitutional monarchy as head of state' and such are a different category than 'parliamentarian democracy' and 'direct democracy'. But how do you combine these categories - and a plethora of others - in a logical way? Not to mention, the same problem with your proposal crops up; what about the non-government categories? Sliders, like in Europa Universalis?

In the end, I'd say your concept is (by far) the best, from what I can think of, but I suspect you'll have to make 'government' a different thing altogether. Let changes in government create anarchy, and use your system for it, while the rest - warfare, labour, what have you - are more like IV's system. Doctrines, values, culture... Yeah, you could link changes in the other categories to non-unit production (labour), unit production / experience (warfare), culture (values such as free speech and universal suffrage), instead of anarchy?

 

[Wessel V1]

I think the mod in general should stick to the original framework. Mods like Rise of Mankind are fun exactly one game for me, which is when I have discovered all content. From there on, it just feels that adding content was the aim in stead of creating a better game. Of course, there are many players who simply want more content, so the mod has (had) a large fanbase. However, I think the beauty of the original RFC was that new features were really scarce and only developed to keep the action going on. Don't underestimate the impassable jungles and swamps and the settler maps so Siberia, Africa, Australia and large parts of the Americas stay empty for a long time. For example, if you look at RFCE for example, it completely lacks action from the 16th century onward; before then there is a bit more but not too much either. In that respect, I think DoC is doing a good job since it adds even more action, without making things overly complicated or making it imbalanced - which was also a huge strength of the original RFC. New techs with new buildings and new units wouldn't do the trick for me. As long as the spirit of DoC to stay close to the game but improve where necessary is maintained, which follows the spirit of RFC in my opinion, correct me if I'm wrong, I'm fully happy with new features. I don't think a steep learning curve is really a problem, if everything is well documented. For example, a list of conditional spawns, the stability maps you're working on etc., those would benefit new players who came from playing RFC. I don't think there will be many new players who haven't played RFC yet anymore, after all CiV is 4 years old already.

 

[VGL]

Wow,i'v just saw the polls again,it seems like many more people want BtS mechanics and content can go if that enables new features and a few say Keep BtS stuff, but RFC mechanics and content are fair game and Keep traditional RFC mechanics, only new content within that framework .

 

[mrrandomplayer]

Wow,i'v just saw the polls again,it seems like many more people want BtS mechanics and content can go if that enables new features and a few say Keep BtS stuff, but RFC mechanics and content are fair game and Keep traditional RFC mechanics, only new content within that framework .

Yeah, I'm one of three or four people who voted for the second option .

 

[VGL]

I'm one of the majority

 

[Hippo8085]

I think Hippo is asking (correct me if I'm wrong) if it would be possible for DoC to take on a SoI style province system

Only for the AI and quite backdoor, because...

incorporate AI settling patterns into provinces, rather than individual tiles. Because of this I think AI settling could become more intuitive and perhaps more spread out. Though I don't know very much about the AI I admit.

It goes something like this:
- Map is divided up into a bunch of city regions, which each contain some important city or two that one wants to see
- Each tile within a region is weighted to a certain value x out of 100, such that in roughly x% of games in which this region is settled, this tile is settled (the actual specifics here may have to be different, but you get the idea).
- Each region and AI pair has some specific weight, there is some number that determines whether the entire region is historical.

The AI settles according to (in order):
- Does the region have a city in it? If yes, don't settle.
- What is the remaining region with the highest weight? Target it. The AI would have to be given some sort of cheat here to ensure that you don't get stuff like the Netherlands settling Cape Horn (maybe they should start with a bit more revealed on their map).
- Settle city within region somewhat randomly according to the individual tile weights. The one city spacing rule still applies.

So, rather than have a settler map for every single civilization, there are 2 maps and a big table.

 

[Leoreth]

It could work. You'd introduce some kind of tab system, so the basic civic screen of IV is merely one tab, called 'government', on which you see the above options. Go to the tab 'warfare', and you see war related options. However, I can't think of any options that can be so detailed as the above 'government'.

Maybe tree wasn't the right word - it's more like branches. You have one civic tree, but branches for different forms of government. So there wouldn't be tabs at all, or really even any explicit tree that you can see, only interdependencies. Whether a civic is a "government" civic doesn't really matter. It's kind of similar how it currently doesn't really matter that a tech is an "engineering" tech mostly (aside of AI and GP purposes). Civics could have lots of dependencies or none at all depending on their context ... which makes them more like fruit in the tree I guess? I forgot where I was going with this.

Something like V's system also doesn't really fit, because that is less about forms of governments and more about concepts attached to one core concept / ideology. And the progression doesn't make a lot of sense.

Not to mention, the same problem with your proposal crops up; what about the non-government categories? Sliders, like in Europa Universalis?

In the end, I'd say your concept is (by far) the best, from what I can think of, but I suspect you'll have to make 'government' a different thing altogether. Let changes in government create anarchy, and use your system for it, while the rest - warfare, labour, what have you - are more like IV's system. Doctrines, values, culture... Yeah, you could link changes in the other categories to non-unit production (labour), unit production / experience (warfare), culture (values such as free speech and universal suffrage), instead of anarchy?

As I said, there doesn't need to be any specific category for stuff to be sorted into, necessarily (although a broad categorization as with techs would be cool). Current civics mostly fall into forms of government (first and second column) and forms of economy (third and fourth column) plus religion and military.

Only for the AI and quite backdoor, because...
It goes something like this:
- Map is divided up into a bunch of city regions, which each contain some important city or two that one wants to see
- Each tile within a region is weighted to a certain value x out of 100, such that in roughly x% of games in which this region is settled, this tile is settled (the actual specifics here may have to be different, but you get the idea).
- Each region and AI pair has some specific weight, there is some number that determines whether the entire region is historical.

The AI settles according to (in order):
- Does the region have a city in it? If yes, don't settle.
- What is the remaining region with the highest weight? Target it. The AI would have to be given some sort of cheat here to ensure that you don't get stuff like the Netherlands settling Cape Horn (maybe they should start with a bit more revealed on their map).
- Settle city within region somewhat randomly according to the individual tile weights. The one city spacing rule still applies.

So, rather than have a settler map for every single civilization, there are 2 maps and a big table.

That's basically scripting AI settlement patterns in its entirety (just with some variance) while completely throwing out its own logic for determining city spots. I don't want to do that much micromanagement to be honest.

 

[Panopticon]

Hey Leoreth, you're doing great.

Some of the most popular Civ IV mods have included major changes or been complete transformations of the game.

I wouldn't fear tinkering with the BtS rules - though part of the charm of the original RFC was that it achieved so much by appearing to do so little more than the core game, which was by the time of BtS a model of good game design(1). Hell, I love Paradox games, but RFC does 80% of the business with 20% of the complexity. DoC, for its part, adds lots of extra civs and wonders to think about, and I still don't understand stability fully, but it stays away from being too complex in other areas like technology or units and has some good innovations(2). Not that these shouldn't ever be changed(3) but that the principles like combined-arms rock-paper-scissors are good and should be conserved. That Antoine de Saint-Exupery tech quote about perfection comes to mind; removals help to counterbalance additions.

Documentation would help. Where, as Leoreth said, are the hands to write it?

I think you should mod what you enjoy. Pursue excellence (including efficiency if you feel it is a worthy goal), treat it as self-expression, listen to the community to hear what they want fixed if anything. Forget also historical simulation and even, for just one moment, realism: The mod is still fun. Keep it that way and it is a great service to many people here.

(1) However, Rhye's code proves that this apparent lightness is an illusion generated by immense and often solitary work.
(2) Though I will never use slaves!
(3) Especially slaves!

 

[Leoreth]

That's a great comment in its entirety, Panopticon.

 

[ImNotHere]

Random question: Is this game multithreaded? I know it's like 10 years old and all, but if it isn't then if it is possible to multithread it, maybe it could be (though I wouldn't doubt it would be a crap-ton of work). One of the only things I dislike about the mod is the slow speed, especially in ~1700 onward, so any speed improvements would be a huge plus for me.

 

[VGL]

So,what's you're plan Leoreth?what changes do you have in mind?

 

[Leoreth]

Random question: Is this game multithreaded? I know it's like 10 years old and all, but if it isn't then if it is possible to multithread it, maybe it could be (though I wouldn't doubt it would be a crap-ton of work). One of the only things I dislike about the mod is the slow speed, especially in ~1700 onward, so any speed improvements would be a huge plus for me.

The game is currently not multi-threaded. Yes, that means that if you have a quadcore processor, you can divide your effective processing power by four in terms of what is actually available for the game. I agree that improvements are needed here, but the fact that we can only mod the DLL and not the EXE is very limited. C2C does some multi-threading that might be worth looking into.

So,what's you're plan Leoreth?what changes do you have in mind?

*mysterious music plays*

 

[VGL]

Will you add more religion,a completely renovated civics and civic related event Mr Leoreth?