Advanced Civ

[f1rpo]

Here's a save - the AIs are willing to trade the second half of biology for 5 bucks.

Thanks. It's a team thing. Apparently, this K-Mod change is responsible (so the bug should also exist in K-Mod):

The trade value of techs is now independent of team size. (ie. the AI will now value techs the same regardless of whether they are giving them to a 2-civ team or a to 1-civ team, whereas previously the tech would be valued more highly for the 2-civ team.)

I could've repaired this by dividing the current research progress by the team size modifier when computing AI trade values, but then techs would've remained cheap to purchase (relative to their research cost) in team games. I think this change was simply ill-conceived and so I've reverted it in v0.97b, which I've just uploaded to the download database.

The tech pace seems a bit fast in your game, with Satellites being researched before 1900. Could be due to the selection of leaders, but probably the TECH_COST_EXTRA_TEAM_MEMBER_MODIFIER (50 as in BtS) also doesn't quite make up for techs being shared within teams. But then, a higher per-member modifier could make Permanent Alliances too unattractive. I guess it would have to be a cost increase based on average team size that affects all players alike. I'll leave it alone unless you think the tech pace is problematic.

By the way, I haven't really play-tested the research penalty for teammates who are researching the same tech; it sounded good on paper ... Is it OK, annoying, interesting? Also (feel free to ignore these inquiries obv.), did the AI start any wars in the first half of the game. Because, last I checked, the AI being reluctant to start wars in team games was a bit of an open issue. (That could be another reason for a fast tech pace come to think of it.)

 

[Lanstro]

We're teching very well this game because of an extraordinarily good position - we got a very large continent to ourselves, and built the Lighthouse, Stonehenge, Great Library, Colossus and all our infrastructure without having to worry about war. The AIs also spent a lot of time warring each other and not having any clear winner. Several of my cities didn't even have a garrison until recently, and you'll see that despite our tech lead, our military strength is still very poor compared to the other teams. This game is far and away an outlier. I did also exploit the 5 dollar thing around 5 times so that of course helped.

Fixing the 5 dollar thing I suspect will ensure there is a greater spread of AI tech positions. Right now the AIs all seem to stick in a pack when it comes to a tech progress, I suspect because they can just buy half-done techs off each other for 5 dollars. I only noticed this behaviour relatively recently, fairly sure in 0.96 the AIs' tech spread was greater.

The team tech research penalty is somewhere between 'ok' and 'interesting'. Most of the time we just tech separately to not eat the penalty, and most of the time it's a no brainer when we should all be contributing to the same tech (eg if someone needs to work their only pigs at the start of the game, so we all hit animal husbandry, or we're looking to unlock a key wonder like Great Library). It's only rare when that call is difficult and interesting - eg whether we should get greedy and hit communism and physics separately, or just concentrate on one to ensure we get one of the great people.

As for AI wars - I'll give a general answer rather than one based on this outlier game. We always play huge map, big n smalls, normal speed, immortal or (lately, emperor), team of 3 vs five other teams of 3 AIs. Have probably played well over 100 starts with this mod. Not sure what you define as an early war, but the earliest wars we've seen have been the AI declare are chariot + archer type rushes, but those are relatively rare. Once you get into axes and swords, wars are quite commonplace, especially if our border cities are encroaching on each other's fat cross.

While I've got the podium, I'll give a bit of feedback on the general rhythm of the game with this mod. In the games we play, setting aside weird outlier games like the one I uploaded or the ones where we do the early warring, the games are playing out a little same-same. We grow to about 5 cities each, rub up against our neighbour AI empires, and then they'll inevitably declare war on us at some point regardless of how peaceful the AI team's personalities are (unless we get lucky and end up with the same religion and so we can hit friendly relations). Much of the time, the game ends right there because we can't handle it. If we do push them back and eventually win the war(s), at around industrial age, one or more of the AI teams decides on domination victory and starts wiping out other teams, which means that unless we already had secured large, very rich lands at this point, we also have to follow suit. This is the stage of the game where the AI takes full advantage of its production bonus, and size 50+ stacks of doom are common, so we either need to have a tech lead or oceans to protect us, or else we're toast.

Essentially this means almost every game plays out like a domination game or a war-fuelled science game. I don't mind if the bulk of games were like this (that's why we've played hundreds of games like this, if not thousands once you include K-Mod), but it'd be nice to have a little variety of the occasional peaceful game that we could get with the base game / PIG mod. It feels like the AI personalities lose a little of their meaning when a team that includes Mali and Ethiopia will still happily war you because your military strength is a little too low.

 

[Cruiser76]

Lanstro, those are interesting observations and pretty similar to my experiences. I didn't play much vanilla so I can't comment on it, but I have played a lot of kmod and advciv.

A fundamental problem is that playing peaceful and/or small is really not an option. If you play small, you just can't keep up in tech or units. I've tinkered with settings etc and it doesn't really matter. A few AIs will expand and conquest like mad and you can't keep up. They will target you eventually. I win every now and then with culture playing small, but only after reloads avoiding a war declaration or securing a defensive pact with a big AI. The AI can play small because it can vassal. Science as small doesn't work. You will never build a spaceship before being destroyed or before an AI does it.

Playing peaceful only works after you have secured a large amount of territory and can keep up with the big AIs. See above. This doesn't even ward off war either most of the time.

 

[kaltruhe]

Very impressive! I'd like to thank you for developing such a nice modmod. I'd like to know that is it possible to use graphical mod like Varietas Delectat with AdvCiv?

 

[f1rpo]

@Cruiser76: If peaceful and small means that you take the land surrounding your capital and never start a war, then it sounds like a pentathlete competing only in three events, the missing events being combat tactics and er... recognizing opportunities for war. OK, I guess you get attacked and prevail (up to a point), so there is combat. Still, it would seem unfair to award victory to someone who "just sat there" – unless they managed their economy exceptionally well or something. I guess, to win, a small civ should have a responsibility to intervene in time against AI rivals that are growing too big. Maybe not feasible on large maps, or for another reason?

Random thought: The difficulty-based tech cost increases in the mod help war mongers in a similar way as Epic speed does. That should only be a significant factor on Emperor+ though.

@Lanstro: Thanks for the detailed reply. Tech pace – makes sense, just an outlier and I wouldn't want every game to have the same tech pace.

Right now the AIs all seem to stick in a pack when it comes to a tech progress, I suspect because they can just buy half-done techs off each other for 5 dollars. I only noticed this behaviour relatively recently, fairly sure in 0.96 the AIs' tech spread was greater.

I also expect that the bug had affected trades between AI civs; though, as far as I can tell, it had existed already in K-Mod. But it seems easily possible that v0.97 changed something to explain the more uniform tech progress.

Most of the time we just tech separately to not eat the penalty, and most of the time it's a no brainer when we should all be contributing to the same tech [...].

Sounds like it's the most that one can get out of that idea.

Early warfare: I've never really tested teams of 3. I've run a game on AI Auto Play now with your settings. One war before AD 1. I like there to be at least one (without a human player having to start it); doesn't feel like Civ when there is no warfare in antiquity. All in all, this one game looked OK in terms of warfare for a watery map like this. Perhaps my tests with 2-player teams had had too few teams. 5 teams of 2 are still just 5 independent actors when it comes to declarations of war. (And when a war does happen, then it's always a big one.)

[...] we've played hundreds of games like this, if not thousands once you include K-Mod

Whoa.

It feels like the AI personalities lose a little of their meaning when a team that includes Mali and Ethiopia will still happily war you because your military strength is a little too low.

ZaraY and MMusa both have only 90% no-war probability at Pleased. Which, of course, still shouldn't make them trigger-happy. Part of their peacefulness comes from having a small military, but that shouldn't be an issue when outnumbering another team on some landmass (or when a third team member provides the manpower). On that note, I would imagine the distribution of teams among landmasses to be a major factor in those games. In the team games I've watched, the AI didn't seem to be quick to take advantage of that; in a human team, I would probably always (advocate to) go after players that are separated from their teammates.
Is it harder to get to Pleased AI attitude in team games? Or do they routinely attack despite being Pleased? Kind of hard to say who is Pleased in a team game. Let's say their members have relations values (+4, +4, +2). That gets rounded to +3 on average and mapped to Cautious as the team attitude. At Cautious, MMusa and ZaraY have just 50% no-war probability. Let's say Suleiman is also on the team, then the vector is (50, 50, 70), average 56. Might be better to feed per-player attitude into the no-war prob table. That would result in (90, 90, 70) in the example if Suleiman has the +2 relations value. (No need to compute an average probability either; my war evaluation code is at the level of players and the war utility values get averaged in the end.)

@kaltruhe: Thanks. There should be no fundamental obstacle. I don't have Varietas Delectat (VD) at hand as a standalone mod, so, looking at a diff between Kat Mod (K-Mod + BAT) and K-Mod, Kat Mod adds some .fpk files, some folders under Art (Effects, GreatPeople, LeaderHeads, Movies, Structures, Terrain, Units). There's a modified XML\Misc\Civ4EffectInfos.xml in Kat Mod; not sure if that's part of VD. Two LSystem files in XML\Buildings and XML\Units\Civ4FormationInfos.xml probably are VD-related. All that stuff is just drag and drop.

XML files that probably need to be merged (because both mods have modified them):
• Units\Civ4UnitInfos.xml: some UnitMeshGroup elements
• Civilizations\Civ4CivilizationInfos.xml: assignment of the new art styles
• Art\Civ4ArtDefines_Unit.xml: a bunch of scale changes

If you or anyone gets it to work, perhaps we can come up with instructions for players who want to do the same thing. Doesn't look like I can make the procedure easier through changes in AdvCiv. Those VD XML changes would cause problems without the rest of VD. Of course a VD-AdvCiv merge could also be uploaded as its own mod; I just wouldn't want to have to keep it up to date.

 

[kaltruhe]

@Cruiser76: If peaceful and small means that you take the land surrounding your capital and never start a war, then it sounds like a pentathlete competing only in three events, the missing events being combat tactics and er... recognizing opportunities for war. OK, I guess you get attacked and prevail (up to a point), so there is combat. Still, it would seem unfair to award victory to someone who "just sat there" – unless they managed their economy exceptionally well or something. I guess, to win, a small civ should have a responsibility to intervene in time against AI rivals that are growing too big. Maybe not feasible on large maps, or for another reason?

Random thought: The difficulty-based tech cost increases in the mod help war mongers in a similar way as Epic speed does. That should only be a significant factor on Emperor+ though.

@Lanstro: Thanks for the detailed reply. Tech pace – makes sense, just an outlier and I wouldn't want every game to have the same tech pace.I also expect that the bug had affected trades between AI civs; though, as far as I can tell, it had existed already in K-Mod. But it seems easily possible that v0.97 changed something to explain the more uniform tech progress.Sounds like it's the most that one can get out of that idea.

Early warfare: I've never really tested teams of 3. I've run a game on AI Auto Play now with your settings. One war before AD 1. I like there to be at least one (without a human player having to start it); doesn't feel like Civ when there is no warfare in antiquity. All in all, this one game looked OK in terms of warfare for a watery map like this. Perhaps my tests with 2-player teams had had too few teams. 5 teams of 2 are still just 5 independent actors when it comes to declarations of war. (And when a war does happen, then it's always a big one.)
Whoa.ZaraY and MMusa both have only 90% no-war probability at Pleased. Which, of course, still shouldn't make them trigger-happy. Part of their peacefulness comes from having a small military, but that shouldn't be an issue when outnumbering another team on some landmass (or when a third team member provides the manpower). On that note, I would imagine the distribution of teams among landmasses to be a major factor in those games. In the team games I've watched, the AI didn't seem to be quick to take advantage of that; in a human team, I would probably always (advocate to) go after players that are separated from their teammates.
Is it harder to get to Pleased AI attitude in team games? Or do they routinely attack despite being Pleased? Kind of hard to say who is Pleased in a team game. Let's say their members have relations values (+4, +4, +2). That gets rounded to +3 on average and mapped to Cautious as the team attitude. At Cautious, MMusa and ZaraY have just 50% no-war probability. Let's say Suleiman is also on the team, then the vector is (50, 50, 70), average 56. Might be better to feed per-player attitude into the no-war prob table. That would result in (90, 90, 70) in the example if Suleiman has the +2 relations value. (No need to compute an average probability either; my war evaluation code is at the level of players and the war utility values get averaged in the end.)

@kaltruhe: Thanks. There should be no fundamental obstacle. I don't have Varietas Delectat (VD) at hand as a standalone mod, so, looking at a diff between Kat Mod (K-Mod + BAT) and K-Mod, Kat Mod adds some .fpk files, some folders under Art (Effects, GreatPeople, LeaderHeads, Movies, Structures, Terrain, Units). There's a modified XML\Misc\Civ4EffectInfos.xml in Kat Mod; not sure if that's part of VD. Two LSystem files in XML\Buildings and XML\Units\Civ4FormationInfos.xml probably are VD-related. All that stuff is just drag and drop.

XML files that probably need to be merged (because both mods have modified them):
• Units\Civ4UnitInfos.xml: some UnitMeshGroup elements
• Civilizations\Civ4CivilizationInfos.xml: assignment of the new art styles
• Art\Civ4ArtDefines_Unit.xml: a bunch of scale changes

If you or anyone gets it to work, perhaps we can come up with instructions for players who want to do the same thing. Doesn't look like I can make the procedure easier through changes in AdvCiv. Those VD XML changes would cause problems without the rest of VD. Of course a VD-AdvCiv merge could also be uploaded as its own mod; I just wouldn't want to have to keep it up to date.

Thank you for quick response.

In summary,

1. Two LSystem files in XML\Buildings and XML\Units\Civ4FormationInfos.xml

Drag from Kat Mod and drop to AdvCiv mod folder

2. Files below in AdvCiv should be replaced by those from Kat Mod.

• Units\Civ4UnitInfos.xml: some UnitMeshGroup elements
• Civilizations\Civ4CivilizationInfos.xml: assignment of the new art styles
• Art\Civ4ArtDefines_Unit.xml: a bunch of scale changes

3. Copy AdvCiv mod folder revised by step 1 and 2 and paste to BAT mod folder

Am I understood right?

 

[Leoreth]

If you or anyone gets it to work, perhaps we can come up with instructions for players who want to do the same thing. Doesn't look like I can make the procedure easier through changes in AdvCiv. Those VD XML changes would cause problems without the rest of VD. Of course a VD-AdvCiv merge could also be uploaded as its own mod; I just wouldn't want to have to keep it up to date.

I have a VD module but it's not really a "module" because it overwrites some files outside of Assets/Modules and still needs to be kept up to date with the base mod in those files it overwrites. It takes the burden of manually merging files from the players, at the cost of me having to do it every once in a while the base mod diverges, and being unusable until I do for some time.

However, thinking about it, most of the problem arises because art style types (units and cities) are different in the base game and most mods derived from it that don't change the graphics, and because LSystem files cannot be modular (which is kinda disappointing imo because so many graphical differences ultimately have to be implemented there). I think what could make a base mod compatible with VD would be to adopt the VD art style types but still have it point to the vanilla art (basically having a lot of redundancy). So for example instead of ARTSTYLE_ASIA for China and Japan they would have ARTSTYLE_CHINA and ARTSTYLE_JAPAN as in VD but still point to the generic Asian art. Then they should be easily overridden by a modular art style definitions file. I don't think this would also work for Lsystem but at least it would get those files to a place where they can be simply replaced.

Is it worth it? It's a lot of work and the diff is not that great if you don't do it. Still I would find it kind of appealing to have an (almost) entirely modular VD for my mod.

 

[Lanstro]

It's not easy to get to pleased in our team games. Vast majority of the time, the AIs will just sit on cautious. Easiest way to get up to pleased / friendly is if you happen to have the same religion.

So one example from a few weeks ago - we spawned next to Justinian, Roosevelt and someone else, and Justinian's early religion spread to us, so we were friendly basically all game. We fought some wars alongside each other too. At around industrial, we were the #1 and #2 teams, with similar sized militaries. This is where I suppose the philosophy of AI who plays to win vs AI that roleplays a little comes into it. We started warring some other team, one or two members of Justinian's team switched into free religion, which I assume meant we slipped off whatever threshold there is that stops them from declaring, and they backstabbed us with devastating effect. Certainly the right move for an AI playing to win, but for variety's sake I would prefer that in a game where we worked pretty hard to maintain friendly status from the start, the relationship would continue on.

I also remember a solo game a few months ago where Mansa declared war on me despite being in reasonably good relations. Even K-Mod Mansa didn't do that. From a variety perspective, I'd like to be able to build smaller militaries when I'm next to the likes of Mansa.

So following on from your response to Cruiser76, I agree that in most games, playing by sitting in a corner and not participating in wars should not be a viable way to victory. However I think that in the occasional game where you start with rich lands next to friendly personalities, cutting corners on military a little to snowball a science lead should be allowed, otherwise every game is some variant of domination. To use your pentathlon analogy, it's perfectly viable to run dead in the religion, wonders or espionage events and still take out the win, so it'd be nice if more than once in a blue moon, skipping military but being awesome at everything else is viable too.

To clarify, where I say we've played hundreds/thousands of games, I mean that many starts. We can often start and fail or abandon 3-4 games in an evening, so the numbers rack up quick. When a game truly gets off the ground it takes about 4 sessions to complete.

Finally, while I'm here, 2 bugs to report for 0.97b. We still can't gift units to each other (the 5 dollar trade / AI tech spread is fixed though). Also in our current game, when we captured Rome's last town, its team's vassal Brennus broke free, which made the score list right above the minimap disappear, and all the other civs no longer showed in the foreign advisor. When we ended turn the game crashed for everyone. Here's the save.

 

[f1rpo]

@kaltruhe: Hard to say what source to use for VD. I don't think that BAT has substantially revised the original VD content, so it's probably better to leave BAT (and Kat Mod) out of it. I've taken a look at the original VD now (turns out it's just 128 MB):
This might be too complicated and laborious if you haven't merged mods before. Anyway, here's what I've gathered:

Spoiler :

* Three FPKs are probably needed; the 4th (BlueMarble) should be optional.
* Only Assets\Art, Assets\Sounds, Assets\XML and Assets\Modules should be needed - but not all of it.
* Art: Effects and Units might suffice. There should be no conflict with AdvCiv's files. Terrain would be only for Modern Roads (which are ugly anyway imo).
* Sounds: Just the Units subfolder I guess.
* XML: GlobalTypes.xml, Units\CIV4FormationInfos.xml, Misc\CIV4EffectInfos.xml, Buildings\Civ4CityLSystem.xml, Buildings\CIV4PlotLSystem.xml, Audio\Audio3DScripts.xml. None of those exist in AdvCiv. Civ4ArtDefines_Unit.xml does exist in AdvCiv, but it's probably OK to just replace the whole file; merging all those scale changes would be a chore. Will have to merge Units\Civ4UnitInfos.xml (mostly UnitMeshGroup changes) and AudioDefines.xml (8 new sound ids starting at SND_A7V_DIE).
* Modules: Looks like the Cultural City styles module doesn't really have to be a module of its own. Could just move its CIV4ArtDefines_Building.xml file into XML\Art and forget about the rest. That leaves the actual "Varietas Delectat" module ...

I'll have to leave it to someone more interested in VD than myself to work that out.

2. Files below in AdvCiv should be replaced by those from Kat Mod [...]

Generally, replacing a whole file will probably not go well. Whatever change karadoc or jdog or I had made to that file would be lost. Instead, the VD additions should be copied line for line into the AdvCiv file (or vice versa). Tools like WinMerge can help with that.

@Leoreth: Thanks. I'm not totally on top of this topic. The idea with the placeholder art styles sounds like it should work. I guess it would make it easier for players to switch VD on and off once they have it installed.

 

[f1rpo]

2 bugs to report for 0.97b. We still can't gift units to each other (the 5 dollar trade / AI tech spread is fixed though). Also in our current game, when we captured Rome's last town, its team's vassal Brennus broke free, which made the score list right above the minimap disappear, and all the other civs no longer showed in the foreign advisor. When we ended turn the game crashed for everyone. Here's the save.

The bugfix for gifting units works for me – it just isn't included in the 0.97b DLL. Everything else seems to be included as advertised. Sorry. Corrected DLL attached. (I don't want to alert people subscribed to the download page just for this one thing.)

Your savegame doesn't crash for me when I end the turn. Perhaps (hopefully I guess) not for you either. So the humans captured Neapolis and Brennus immediately broke free. That shouldn't have happened because Sitting Bull on the master team was still alive. I can't tell from looking at the code what went wrong exactly. Whatever error it was, it may well have also caused the subsequent crash. So a savegame from before the conquest of Neapolis would be helpful. Otherwise, maybe I can reproduce it in a single-player game (easier to debug) if I produce a game state where a team of two has a vassal; might be all that it takes.

I'll reply about the AI/ balance issue later.

 

[f1rpo]

I agree that in most games, playing by sitting in a corner and not participating in wars should not be a viable way to victory. However I think that in the occasional game where you start with rich lands next to friendly personalities, cutting corners on military a little to snowball a science lead should be allowed, otherwise every game is some variant of domination.

Generally, I'd like the game to reward players for adjusting their priorities over time based on the diplomatic situation, available technologies, any kind of opportunities. I don't see why every war ought to lead down the road to defeating all major rivals militarily. Especially for dogpile wars (I mean any kind of joint war really), veritable stacks of doom shouldn't be needed in order to make a difference.

It's not easy to get to pleased in our team games. Vast majority of the time, the AIs will just sit on cautious. Easiest way to get up to pleased / friendly is if you happen to have the same religion.

Friendly is a bit of an open issue, but Pleased – shouldn't be easy exactly or a matter of course, but it sounds like there might be a team-related issue. Open Borders perhaps. Let's say, if teams A and B are lined up geographically on an oblong continent:
A1-A2-A3-B3-B2-B1
Then, A3 and B3, being close to each other, might soon reach +2, while all other pairs might still be at +1 or 0. The point being, I guess, that +1.5 or so relations (on average) from OB may require all pairs to be close while, for warfare between the teams, one pair being close suffices. Just speculation. I'll try to look at some examples via AI Auto Play.

So one example from a few weeks ago - we spawned next to Justinian, Roosevelt and someone else, and Justinian's early religion spread to us, so we were friendly basically all game. We fought some wars alongside each other too. At around industrial, we were the #1 and #2 teams, with similar sized militaries. This is where I suppose the philosophy of AI who plays to win vs AI that roleplays a little comes into it. We started warring some other team, one or two members of Justinian's team switched into free religion, which I assume meant we slipped off whatever threshold there is that stops them from declaring, and they backstabbed us with devastating effect. Certainly the right move for an AI playing to win, but for variety's sake I would prefer that in a game where we worked pretty hard to maintain friendly status from the start, the relationship would continue on.

Was it too sudden? Obviously, if their team switched from out of nowhere and immediately attacked, that's a problem. If It's foreseeable, then the civics change could be an interesting twist, forcing you to abandon that other war (foregoing Domination this once ) in order to deter Justinian's team. The impact of attitude on AI declarations of war seems to be extra murky in team games. As I wrote before, the unnecessary averaging of attitude levels can make the AI more unscrupulous than one would expect. I agree that it should be possible to have friendly relations in the late game when religions fall by the wayside.

I also remember a solo game a few months ago where Mansa declared war on me despite being in reasonably good relations. Even K-Mod Mansa didn't do that. From a variety perspective, I'd like to be able to build smaller militaries when I'm next to the likes of Mansa.

In a situation like this, I'd really need to look at the UWAI log file to see why MM declared war. That can be enabled in GlobalDefines_devel.xml, but I can also get it from a savegame prior to the DoW (ideally at/ around the time when war preparations begin). And having that savegame is of course helpful in any case.

 

[crullerdonut]

I'm fulfilling my promise in the K-mod thread to try out this mod. It keeps crashing when I end my turn. I use the CD version of the game, patched to 3.19, and I'm using AdvCiv 0.97b.

 

[Lanstro]

Haven't managed to get a game together with my group lately, but here's a save from a few minutes before the save I uploaded that might help you recreate the foreign affairs bug we had.

 

[f1rpo]

@crullerdonut: Sorry about that. A crash in such an early run-of-the-mill game state really shouldn't happen. I'm attaching a hotfix. (To be unpacked into the Assets folder.)

@Lanstro: Thanks; that's just the right moment. Still no crash or anything, and this is consistent with my hypothesis that a dead player had been cached among the living vassal players. That cache gets recomputed upon loading a savegame. I'll assume that this problem is fixed in my local copy and will include the fix in v0.97c.

 

[crullerdonut]

@crullerdonut: Sorry about that. A crash in such an early run-of-the-mill game state really shouldn't happen. I'm attaching a hotfix. (To be unpacked into the Assets folder.)

@Lanstro: Thanks; that's just the right moment. Still no crash or anything, and this is consistent with my hypothesis that a dead player had been cached among the living vassal players. That cache gets recomputed upon loading a savegame. I'll assume that this problem is fixed in my local copy and will include the fix in v0.97c.

Your hot-fix worked! I've played some more turns, and it hasn't crashed yet.
If you don't mind my asking, what was the cause?

 

[f1rpo]

Since you've played a few more turns, it's probably not a spoiler: Git commit
There's a sanity check I've added to stop the AI from demanding items of negligible value in tribute. A guard was missing for the case when no item was found. I've probably tweaked the code somehow for v0.97 or moved things around; I doubt that this had been broken prior to v0.97. Too easy to run into.

 

[crullerdonut]

Since you've played a few more turns, it's probably not a spoiler: Git commit
There's a sanity check I've added to stop the AI from demanding items of negligible value in tribute. A guard was missing for the case when no item was found. I've probably tweaked the code somehow for v0.97 or moved things around; I doubt that this had been broken prior to v0.97. Too easy to run into.

With the hot-fix, nobody demands anything when I go to the next turn without it crashing. So, don't worry about a spoiler, since I don't know who it was who demanded something from me. (Although, Wang Kon demands Civil Service from me the moment I research it, so maybe it was that guy)

I'm impressed that you were able to debug the problem that quickly. Just a missing pair of curly brackets, eh? No wonder it was easy to miss.

I have run across some things that I'm confused about.
Firstly, visible in screenshot 0014: I don't remember the vanilla BTS map generator ever putting 5 fish in that close of a cluster. Not saying that it's a K-mod- or AdvCiv-only thing, but dang, that's a lot of fish.
And then my next question: how did Napoleon get over to that island? (Screenshot 0018). He doesn't have Optics or even Compass, and the island is inaccessible with a Galleon (I checked by giving myself some with WorldBuilder). EDIT: I included the earliest save that I had, and Napoleon is already there.
And then finally, what are all those red circles on the map in WorldBuilder? Just wondering.
Edit: I think that the red circles are spots where the computer thinks that a better improvement can go there (the glowing button), and the blue circles might be places where it thinks are good spots to found a city.

 

[Cruiser76]

Napoleon got there from Lyons. One of the cultural squares is adjacent to a coastal square of the island.

 

[Leoreth]

Napoleon is quite good at getting to and from islands, as long as they are not too remote.

 

[crullerdonut]

Napoleon got there from Lyons. One of the cultural squares is adjacent to a coastal square of the island.

Oh you're right. When I put a Galley down in WorldBuilder, it didn't work because it wasn't within my own cultural border. But Galleys can go across oceans, so long as they are within their cultural borders.

I found another issue with Advanced Civ, though. The majority of Space Ship parts are not necessary, as seen in Screenshot 0019. Notice that for the Cockpit, Life Support, Stasis Chamber, and Docking Bay, the required number is: 0. So, you can launch with a single Thruster, a single Engine, and a single Casing, and that's it.
The correct requirements are shown in screenshot 0020.