1. With two AI at war, both may have diplo attitude on asking to make peace "We'd love to but you will have to contact them" - have not seen that before, does not make a lot of sense. Perhaps nothing to do with your mod, idk.
That probably is due to K-Mod, but I don't think it's a bug. In the unmodded game, AI's could always be bribed into peace as long as they weren't losing very badly. In K-Mod, the same rules apply but with an extra condition: you'll get that "talk to them" response if the third civ refuses to talk to the civ you're talking to. ie. previously it was possible to broker peace deals between two civs who refused to talk to one another, but in K-Mod that is no longer possible.
Incidentally, if either of the warring civs is a human player then both will have the "talk to them" response - and that's the same in the unmodded game. I suppose when that happens, at least one of the leaders must be lying...
2. Caesar says "Senatus Populusque Roma saluto vos" - this is incorrect - Senatus Populusque Roma = Senate and People of Rome [plural] and "saluto vos" = I salute you [singular]. So should be "vos salutamus" = [we] salute you or "vos salutant" [plural]. Not a game-play point
Thanks. I mentioned it to the author of ActualQuotes, and he says you're probably right. Neither of us speak Latin, so we'll just trust you on that.
I think I'll use the second one:
"Senatus Populusque Roma vos salutant".
There is one serious defect in the diplomacy, again nothing to do with your mod, when it comes to giving units. [...]
Yeah. That is a significant missing part of diplomacy. I agree that it could use some work.
I think the main issue is that gifting units is a somewhat subtle act, and it's not easy for the AI to tell how much it actually helps them. For example, if you gift the AI a bunch of old units that you would otherwise have just deleted; then you've gotten rid of something you didn't want anyway, and the AI will probably spend a much of money upgrading the units, and it might not have wanted those units anyway... So in that kind of situation, the AI probably shouldn't be too grateful for the gift. It may have actually done them more harm than good. On the other hand, you gave some examples where the gift of units was probably very valuable...
It's not easy for the AI to tell the difference between these things; and currently it doesn't even try. The main risk of making the AI try to evaluate this stuff properly is that it may leave the AI valuable to exploitation - where if a human player knows what the AI will be grateful for, the human player might work out how to
game the system to get lots of diplomacy points for gifts that really aren't worth anything.
But I think we can probably get something that's a bit better than the current system without risking any significant exploits. I've got some ideas... basically I'm thinking that the AI should only value the gifts if they think they'll need them. They shouldn't care about gifted units if they are not at war and aren't in a start of 'alert', or if they are already winning their wars comfortably. They're should only take the units if they can afford to support them; and they should be more grateful for more powerful units; and more grateful if they are losing a war badly; and they should check that the units are actually a useful type of unit, and in a useful area, and so on. All these things will need to be evaluated somehow. -- I'll see what I can come up with.
Another problem is that the AI might not know what to do with the units that you give to it; because usually the AI assigns a role to each unit when it first starts building it. eg. city defenders, city attackers, defensive collateral, etc. The AI isn't too bad at deciding which units they need to build, but they don't currently have any way of deciding which role an existing unit should take up. So if you give the AI a bunch of riflemen, then they'll all end up using the default role for riflemen - which is city defence, regardless of what promotions the riflemen have... -- anyway, that's a secondary issue. My point is just that giving units to the AI might not always be as useful to them as you expect.
[...] So this latest version seems to be less aggressive or just easier? My diplo is good, not sure it is that good ...
It's hard to say... The way the AI decides
when to go to war is still essentially unchanged from the original AI - but K-Mod has a completely new way of evaluating
who to go to war with. The the
when and
who are obvious somewhat related, and so the result is that K-Mod may turn out to be slightly less aggressive because sometimes my new AI will say "there are no sensible targets for war" when the original AI would have basically just attacked anyone. -- I'm not really sure how often that happens, but I think I'll make some adjustments to make the AI a bit less scared of war if the game is close to ending.
--
@karadoc
Can you explain how the AI decides if and when to expand?
They don't think about it very much. Basically they just look for potential city places, and if they can see any then they consider building a settler - and if they build the settler then the settler will walk to one of the city places and make a city.
It's true occasionally one of the AI's fails to expand properly; but I don't think there is any single problem that causes it. Sometimes it's because their only good city starting unwisely building an early wonder than took way too long... and sometimes it's because something went wrong in their economy to convince them that they can't afford any more cities; and sometimes it's because they are too busy building units for war, etc.
There is actually quite a lot of randomness used in the decision making for what the AI builds; and sometimes they just get 'unlucky'. Another way of looking at it is that they sometimes just make a mistake - as human players do. -- In general I've tried to make the AI not make obvious mistakes; and I certainly could tell them "don't ever even consider building a wonder if you still need more cities", but I'm not sure I want to do that because I don't want them to become too rigid and predicable. I want the AI to be as smart as possible, but also varied and unpredictable.
That said, there are some definite problems that I know about which can slow down the AI, but which I haven't gotten around to fixing. The wonder building thing is one example there are still some parts of the original AI which can cause them to start building a wonder and those parts are... not very good. Another example that I've been half-intending to fix is that if the AI's capital happens to have a lot of seafood, they often spend way too long building work boats. They want to get workboats on all the seafood even if they can't actually grow their city enough to use all that food...
Also:
Is the AI aware of the "raging barbarians" setting? Does it adapt?
How does barbarian AI work? Does it have any goals?
The barbarians don't really have any goals. Or rather, I guess you could say their goal is to destroy all civs, but they are very disorganized. The barbarians and deliberately reckless. They'll often attack even when they know they'll probably lose. They can't build cultural buildings or settlers, and so they can't expand.
AI civs are not conscious of the "raging barbarians" setting. They wont' make any special preparations for it; and so they may struggle with it a bit more than human players do. But the AI is generally pretty cautious with its city defenses anyway...
As I march in to the Russian capital and only Russian city, I notice that every turn 2 more archers are appearing and pop is going down by 1 every turn. (3 then 5, then 7). Am I missing something?! I can't pop-rush 2 archers per turn for multiple turns so how can the AI?! I nearly pummeled my LCD screen in frustration....
[...] Save is attached...please fix
In the save you posted, you not at war with Russia, and Russia is not building archers. Do you have a save where what you've described is actually happening? It's hard to say without seeing it happen, but I suspect it is not a bug.
As noto2 points out, when the city's overflow productivity is full, the overflow beyond the usual overflow spills onto the next unit (whereas in the unmodded game, this excess overflow was actually silently turned into gold). It's possible to build multiple units in a single turn. Also, because of this overflow system, it's possible to whip for full value multiple turns in a row. Here's what I mean: as you probably know, if you try to whip something that you haven't started building then you only get 2/3 of the usual hammers per pop from the whip. But because of this overflow system, the excess overflow might 'start building' a new unit on the same that the previous unit was built.
Anyway, even with all that the city would have to have pretty good natural productivity to be able to get 2 archers per turn for multiple turns in a row. It's certainly possible though.
By the way, this multi-build functionality was really meant to fix late-game cities from being 'maxed out' in productivity with game speeds faster than 'epic'; and also to increase the viability of cruise missiles - which would usually be too cheap to build without wasting productivity. The fact that the multi-build stuff sometimes has a significant effect on the early game is a bit unintended. In particular, the chain-whipping thing I described is an unintended side-effect, and I don't really like it but there isn't really anything I can do about it without making up new and obscure rules, or reverting the changes, so I'm just going to leave it alone.
BTW...what mods do incorporate K-mod?
Not many. Basically just BUG / BULL with some missing bits, and some added / changed bits, and ActualQuotes. -- The "Better AI" mod was the starting point for the AI in K-Mod, but the bulk of that AI has been rewritten now anyway. I also used someone's global warming mod at some stage, but it didn't work properly so I made my own completely new system for that - but I think there are still some residual bits of code from the other global warming mod left in K-Mod; xml parameter names and stuff like that.
. (Clue: they had two-thirds of France before it, and hardly any after). They won some battles, and yes at Agincourt longbows were pretty decisive.
... this isn't exactly a high priority though. Forest preserves are already situationally useful for their +1 
-- The answer to that question is... well basically I don't know. I know that a few people from different mods have tried to use some or all of K-Mod's stuff, but I don't really know which mods if any have been keeping up-to-date with K-Mod changes. (Apart from Mattygerst's mod. Strangely enough, that's actually slightly ahead of K-Mod at the moment.)
