Making Cradle 3+ fully compatible with the Apolyton Edition

[Kull]

Well, the easiest way to deal with that is to toggle "Pollution Off" from the Rules screen when starting a new campaign (see attached). If you still want to play with "Pollution On" and have a situation where pollution can increase for some period after which something will drive it down to zero, that is something you would have to mod into a game. Fortunately the mechanics aren't that difficult to understand. They use the same mechanism which reduces population pollution - an attribute assigned to one or more buildings in the "buildings.txt" file. As an example, let's look at the "Aqueduct" in Cradle:

IMPROVE_AQUEDUCT {
DefaultIcon ICON_IMPROVE_AQUEDUCT
Description DESCRIPTION_IMPROVE_AQUEDUCT
EnableAdvance ADVANCE_AQUEDUCTS

PrerequisiteBuilding IMPROVE_GRANARY

ProductionCost 1200
Upkeep 3
FoodPercent 0.15
StarvationProtection 5
PopulationPollutionPercent -0.1
ProductionPollutionPercent -0.4
}

In reality the Aqueduct doesn't affect production pollution, but I added that to this example so you know it exists. Anyway, you can add those attributes to ANY building, and in ANY percentage. Also, the AI will construct EVERY building, so eventually these will appear in every AI city, and do their reduction "thing". So you'll get the "across the board" result you are looking for. The AI won't "know" it is doing good things to abate pollution, it will just happen.

The design part would be if you assign these attributes in different values to structures which become available at different times, so you can have the effect of "pollution grows and then at some point begins to decrease". For example, if 7 structures each reduced population pollution by 20%, eventually they would all be built and the net effect would be a city-by-city process in which the they produce zero population pollution and THEN provide a 40% decrease (into the environment).

Best of all - if you have an ongoing campaign and fear that pollution is about to destroy it, you can make these adjustments in "buildings,txt" and the next time you load a save game, they will take effect immediately! For example, I changed the PopulationPollutionPercent in Aqueduct to -1.1 and it took effect right away.

[sevenfm]

I think I've read somewhere that AE supports kind of civ3-like strategic resources, when you can only build certain units/building if you have that resource. Do you plan to implement something like this in this mod?

[Kull]

I think I've read somewhere that AE supports kind of civ3-like strategic resources, when you can only build certain units/building if you have that resource. Do you plan to implement something like this in this mod?

There will be at least two. The "Horse" good is a requirement for building most cavalry units, and you cannot build "Nukes" without the "Uranium" good. I'm reluctant to go much beyond that, since it remains to be seen how well the AI can handle this. For example, requiring "Copper" or "Iron" as a requirement for the various infantry units would probably just strangle the AI.

I'm still looking at the possibility of "chaining" TIMPS, specifically farm-type improvements. For example you could build them on Rivers, while going inland from there requires a neighboring farm (replicates the need for irrigation canals to extend out from water sources). The problem is there aren't enough rivers in CtP2, and the AI is simply clueless about the mechanics involved. Possibly it could work if Hills could be sources (simulating run-off) or some sort of very common "Well Good" was made available. Probably won't happen, though.

[sevenfm]

I really like the idea of strategic resources, and in CTP2 regular resources don't look really important, I'm also concerned of how AI may use them, for example in Civ3 AI is really obsessed with gaining resources and will build cities on them (even before they are known) and fight for them, probably not anything like that in Ctp2 even with AE.

[Kull]

It was a design difference between the two games. The CtP2 AI looks at them primarily as trade goods, whereas Civ3 added coding for the AI to see strategic value. That doesn't exist in CtP2, so while you can create strategic resources now (thanks to the Source Code project), the AI doesn't recognize them as such. In the case of Horses, they are fairly common, and - most importantly - they exist on the terrain-types the AI likes to settle on. So the odds are good that most civs will have horses or will trade for them. By contrast Copper and Iron exist in mountains, and - especially in the early game - the AI isn't likely to have those resources sitting within the active work zones of the smaller early cities.

Could the availability percentages and source terrains be redesigned so as to widen the availability of additional strategic-type Goods? Probably so, but it's not something I'm looking at.

[Cyberguy]

Well, the easiest way to deal with that is to toggle "Pollution Off" from the Rules screen when starting a new campaign (see attached). If you still want to play with "Pollution On" and have a situation where pollution can increase for some period after which something will drive it down to zero, that is something you would have to mod into a game. Fortunately the mechanics aren't that difficult to understand. They use the same mechanism which reduces population pollution - an attribute assigned to one or more buildings in the "buildings.txt" file. As an example, let's look at the "Aqueduct" in Cradle:

IMPROVE_AQUEDUCT {
DefaultIcon ICON_IMPROVE_AQUEDUCT
Description DESCRIPTION_IMPROVE_AQUEDUCT
EnableAdvance ADVANCE_AQUEDUCTS

PrerequisiteBuilding IMPROVE_GRANARY

ProductionCost 1200
Upkeep 3
FoodPercent 0.15
StarvationProtection 5
PopulationPollutionPercent -0.1
ProductionPollutionPercent -0.4
}

In reality the Aqueduct doesn't affect production pollution, but I added that to this example so you know it exists. Anyway, you can add those attributes to ANY building, and in ANY percentage. Also, the AI will construct EVERY building, so eventually these will appear in every AI city, and do their reduction "thing". So you'll get the "across the board" result you are looking for. The AI won't "know" it is doing good things to abate pollution, it will just happen.

The design part would be if you assign these attributes in different values to structures which become available at different times, so you can have the effect of "pollution grows and then at some point begins to decrease". For example, if 7 structures each reduced population pollution by 20%, eventually they would all be built and the net effect would be a city-by-city process in which the they produce zero population pollution and THEN provide a 40% decrease (into the environment).

Best of all - if you have an ongoing campaign and fear that pollution is about to destroy it, you can make these adjustments in "buildings,txt" and the next time you load a save game, they will take effect immediately! For example, I changed the PopulationPollutionPercent in Aqueduct to -1.1 and it took effect right away.

Thanks for the info, I am aware I can turn pollution of at the start of a game but i like that's its there, just not that it becomes the overarching dynamic to play too because of the impeding doom scenario and I don't like seeing my lands inundated.

[Kull]

Siege Attack Redux: Originally the far future "Bio-Terror" attack, this was reconfigured into an Ancient Era special attack and was extensively reviewed in Post #127. First the good news. As playtesting has shown, the AI can and WILL deploy this attack against the human player. In the attached example I altered the effect to -10 unhappiness (and also noticed the name of the attack needed to be changed in the "exp_str.txt" file), and you can see the result. Huge levels of unhappiness that can't be managed at the city level and require an Empire-level happiness boost to offset. And the building of the Nomad was locked at 14 for the next five turns. Sooooo, it works right?

[Kull]

Siege Attack Redux (continued): Well...yes and no. In playtesting, the AI used the attack several times against AI opponents, which was a good indicator that the attack was working as intended. Not much use spending time and effort to make this thing look good if the AI won't use it, amirite? Unfortunately, that highlighted a problem. A big problem.

In the example below, a Carthaginian Wonder Unit deployed a Siege Attack against the Assyrian city of Arbela. You know the attack is successful, because the red "Siege Symbol" can be seen on the Arbela City Status Bar (failed attacks do not display this icon). However, the "Happiness" window does NOT show any effect from this attack, and the Happiness level itself is still quite high (78). At the time of the attack, Arbela was building a Slinger (top picture) and - even though all production is supposed to halt for 5 turns - we can see (bottom picture) that the city has completed the unit and started work on a Shrine. Meanwhile the happiness level remains solid at 78.

Unfortunately this is NOT an outlier. I tested it again and again and again, altering settings every time, and got the same result every time. Siege attacks deployed against AI civs, whether by the human player or another AI, have NO EFFECT. Really disappointing. That said, I will keep the attack in Cradle 3+, because it is one of those rare instances where the AI has an in-game advantage over the human player. In other words, given half a chance, the AI WILL deploy this attack against you, so....player beware!

[sevenfm]

I wonder if it's possible to use script to intercept this action and apply some effects, I think AoM4 had siege implemented in that way, so that under siege units in city cannot be healed.

[Kull]

Unfortunately I'm not particularly good at scripting. In a few cases I've been able to spot problems and make fixes, and even create some new ones - but only when I can copy existing scripts which do almost exactly the same thing. I wouldn't even know where to begin when it comes to replicating a faulty game mechanism like this.

On the plus side - and I'll get into this a bit more at some point - I've been carefully tweaking the AI and it currently uses EVERY ancient era special attack in it's arsenal, which is light years ahead of the situation when I started on this project.

[Kull]

Special Attacks - "Era Dependency" Fix:

ISSUE: I've talked about this throughout the individual Special Attack reviews (and more specifically in Post #111), but let's go through it once more. In Cradle, many of the special attacks were repurposed so they could apply in the Ancient Era. The idea being that - for example - a Modern Era legal attack (Injunction) which stops production for one turn could also be used in the Ancient Era as a form of sabotage. The effects of the attack are IDENTICAL, but the graphics, sounds, and messages were all changed to reflect the Ancient Era version. The problem of course is that eventually the Modern Era arrives and Lawyer units appear, but instead of filing Injunctions or issuing Lawsuits, they are setting fires and assassinating civilians.

SOLUTION: I'll go into more detail below, but the solution involves saving the game at a specific point and then copying in a "Swap Folder" which contains a revised set of files. Specifically, when the player gains the "Age of Reason" Advance, that is considered the beginning of the "Modern Era" for file swap purposes. These new files contain everything necessary in order to change all the graphics, sounds, and text from the Ancient to the Modern Era. I've tested this and it turns out that you can swap almost any file and see the changes immediately simply by loading your Save game. No other actions are necessary. The exceptions to that are slic programming files (anything with an .slc suffix) and severl of the "string" files, "info_str.txt" in particular (the file which holds the message text). The former requires the player to use the /reloadslic command, but I wanted to avoid even that, so there are ZERO slic files included in the Swap folder. As for info_str.txt, the problem is that even a ONE WORD change to a SINGLE MESSAGE results in the corruption of EVERY MESSAGE! Accordingly, that file is also not included. Fortunately there were only 4 messages which had a strong Ancient-Modern dichotomy, but even with those I was able to craft generic alternatives which work in any era.

PRELIMINARY FIXES: To reduce the number of File Swap changes, I was able to create alternative Great Library entries, so that BOTH "Sabotage" and "Injoin" AND "Assassinate Civilian" and "Sue" appear in the GL at the same time, regardless of era. In similar fashion, all required sound files and almost all graphics are already present in the appropriate game directory. Which means all that's necessary is to swap the text files which link to them. Apologies if that sounds a bit esoteric, but trust me - it reduced the number of required file swaps by a factor of ten!

ERA-DEPENDENT SPECIAL ATTACKS: Not all special attacks have this problem, but here are those which do:
* Sabotage/Injoin: Almost everything associated with this attack has to change. To include the sound and ALL the graphics (button, cursor, messageicon, sprite & city status bar icon)
* Assasinate Civilian/Sue: Same with this attack. Have to change the sound and all the graphics (button, cursor, messageicon & sprite)
* Soothsay/Advertise: The change affects only the sound and the sprite
* Indulgences/FaithHeal: A relatively limited change which affects the sound, button & cursor
* Siege/Plague: Actually these are completely different attacks, but they share the same sound and sprite. Since the first is primarily Ancient while the second is Modern, we need a sound & sprite swap.
* Assasinate Ruler/Bomb Cabinet: The name of the attack stays the same, but the graphics & sound will take advantage of the File Swap process to shift from "knife" to "bomb" at the start of the modern Era.

UNIT CHANGES: A small number of units are able to deploy ancient era special attacks and yet still exist in the time period after the "Age of Reason". However, there is a solution which prevents them from deploying inappropriate special attacks from that point forward
* Cleric: A revised "units.txt" file removes his ability to Sell Indulgences (which otherwise will feature "FaithHeal graphics and sounds). There remains a small disconnect since this unit will retain the "Soothsay" attack, but the only difference is that you'll now see and hear the Advertise Sprite and Sound. That's not optimal, but the problem eventually goes away when "Cleric" upgrades to Televangelist (who does not have Soothsay).
* Patriarch: A revised "units.txt" file removes his ability to Sell Indulgences (which otherwise will feature "FaithHeal graphics and sounds). Does not have the Soothsay attack, so that's not an issue.
* Poisoner: A revised "units.txt" file removes her ability to Sabotage and Assasinate Civilians since those become purely legal special attacks. She still retains the ability to Assasinate Ruler (now with a bomb) and is the only unit with that capability. It's also possible that the AI will use this unit more if that's the only thing she can do.
* Wonder Units: Actually NONE of their abilities will change, however just be aware that the sound and sprite for the "Siege" attack will shift over to "Plague". All other graphics remain the same.

FILE SWAP MESSAGING: This does not refer to the special attack messages themselves (as noted above, they cannot be altered), but rather the messages and visual cues which tell the player which File Set is active and when it's time to make the change.
1) The Main Menu will have a graphical indicator telling the player that files from either the Ancient or Modern Era are active (see attachment - 1st pic)
2) There is a message which pops up on T2 and tells the player that yes, you have the correct Ancient Era files or no, you are using the Modern Era ones and need to swap back. Since the "as-delivered" mod will start with the Ancient Era files already active, the latter issue would only occur when a previous game had moved into the Modern era and files were swapped, after which the player decides to start a new Ancient Era campaign. This message includes a messageicon which also tells you which swap file is active (see attachment - 2nd pic)
3) There is a second pop-up message which plays after "Age Of Reason" is discovered, telling the player that it's time to do a file swap.
4) A new GL Concept entry describes the Swap File process, and each of the new messages has a "Library" button which pops open a window containing this entry (see attachment - 3rd pic)

FILE SWAP PROCESS: As the above section clearly indicates, although we've been talking about File Swaps which shift certain sounds and graphics from the Ancient to the Modern Era, the player ALSO needs to be able to transition from Modern to Ancient (for example, when starting a new campaign). Performing this task is extremely easy. The Cradle 3+ directory contains two self extracting archives that will automatically copy the swap files into the correct directories. For the Ancient Era, just doubleclick the file called "Ancient.exe" (see attachment - 4th pic) and click the "Extract" button. And that's it. The Modern Era file swap works exactly the same, except instead you will doubleclick the file called "Modern.exe"

In closing, this process is ENTIRELY OPTIONAL. It is typically performed only once per campaign, can be done in any year, and - most importantly - is not necessary in order to play Cradle 3+. Your game will not crash or fail in some fashion, but the "immersion factor" will be reduced if you are using the wrong set of files for your era.

[Kull]

CTDs - Final Resolution:

As first reported in Post #88 sometime around the middle of August I began to encounter CTDs, and as reported a few posts later, was able to trace them back to the AI control files in the aidata folder. The problem was temporarily resolved by restoring most of those files to the Cradle defaults, but that was just a temporary fix since the Source Code team made a number of improvements to the AI, much of it reflected in the AE versions of those files.

Since then I've been engaged in the now-completed "Special Attacks" project, but several weeks ago I returned the focus to the CTD problem. I'll spare you the tedious details, but this time I went through every file and didn't make any changes without confirming linkages to the other AI files. And the result was good. Cradle 3+ now has a full set of altered aidata files which make use of the new SC team changes and....the game does not CTD. Three test games ran down to 500 BC, another went to 1 AD (see attached), and not a hint of an issue.

Equally important, the AI is now making full use of ALL the Ancient Era Special Attacks:
* AI Diplomats establish embassies, hold receptions, investigate cities AND Hear Gossip.
* All the AI civs build Slavers and use them to Raid cities and capture free roaming Nomads.
* AI Religious units Convert cities, Sell Indulgences and Soothsay
* AI Assassins use both Sabotage AND Assassinate Ruler
* AI Spies investigate cities, steal technology, and instigate successful revolts
* And yes - ignoring the problematic aspects - even Wonder Units will deploy the Siege attack

[Kull]

AI Gameplay & Frenzy:

Eliminating CTDs is obviously important, but that couldn't be the "end state" for the AI. I spent a lot of time tweaking the unit build lists, AI personalities (more on this later), and AI priorities so the end result is that all civs are NOT carbon copies of one another. Inevitably there's a lot of visible commonality since all civs must choose from the same pool of unit sprites, but now some civs are more focused on religious units, others on diplomats, some more willing to explore than others, etc.

The one area where I couldn't move the needle involved AI conquest. AI civs certainly go to war with each other (more often than not) and are fully capable of invading enemy territory and waging stack-on-stack combat and pillaging. But cities rarely changed hands and there didn't seem to be much I could do to change that.

So, reluctantly, I took a look at the frenzy.slc file. Again I'll spare you the details, but eventually I settled on a modified version of "APOL_frenzy.slc" since the base file was presumably created to be at least somewhat compatible with the AE version of CtP2. But regardless of which file I chose, ALL frenzy files generate an enormous number of circular errors, and the only way to stop them is to disable the Debugslic setting in userprofile.txt. That is completely sub-optimal for mod development, but there was literally no alternative. Worth noting that Frenzy has the exact same errors in a "standard" AE game, so it's not something unique to Cradle - the frenzy code just doesn't play well with AE.

That said, eventually I was able to develop a working version of Frenzy (still requires "Debugslic=No") and played a number of campaigns. Enough to be quite confident in saying that while "Frenzy" does appear to make the AI more aggressive (some cities do change hands, albeit no more than two per playtest), that comes with FOUR downsides:

1) The biggest issue is that game turns start to take MUCH longer to complete. Having two minute "end turns" starting around 1800 BC is really awful, and it only gets worse. Admittedly it might not seem so bad during a normal game, but it makes playtesting EXTREMELY time consuming.

2) The Frenzy code inserts FAR more civilian units into military stacks than you would see otherwise, and that reduces the number of special attacks one might otherwise see per turn. You also don't see many Diplomats, Slavers, or Prophets traveling far from the homeland. They go after neighboring civs, but that's about it.

3) AI Map Exploration seems to be hindered. Ships continue to explore, but not as many or as far, and most are unoccupied. Meanwhile, Land Units rarely move too far from the homeland & border areas. Probably because the frenzy code is grabbing them and pulling them into stacks.

4) The AI is extremely aggressive with ALL civs, so embassies with other civs are rare, and they go to war with all their neighbors. It makes for a much less interesting game.

It's possible these results are driven at least in part because the "human" player is the Barbarians, and frenzy is primarily intended to go after a normal Human civ. Even so, I don't like a lot of what I'm seeing, especially since Martin Guhmann put a lot of effort into fixing the AI and explicitly said that Frenzy isn't needed with the SC fixes. In addition, it wouldn't be surprising if Frenzy and the SC AI are fighting each other, and thus making both of them less effective.

Another factor is that most of my games involve large continent size islands, and historically the CtP2 AI works better when cross-water algorithms are minimized. Potentially therefore, "more land" would help Martin's improved AI to function better. In the attached example, we see that Babylon controls most of the northern continent, and while Egypt was able to colonize the southern shore, they have already lost Amarna and would probably lose more cities if the playtest had continued.

*** FINDING: At this point, I'm pretty sure the modified "Frenzy" code is not going to cause CTDs, so I'll keep it as a deliverable with Cradle 3+, but will NOT activate it in the default game settings. However, players that want Frenzy can easily activate it themselves by editing a single line in "script.slc".

[Kull]

Frenzy Revisited: After taking another look at the Frenzy file, I was able to identify the section of code which causes the slic errors (lines 244-250). It's supposed to identify the civilian units which should not be pulled into stacks, but we know from the previous playtests that the errors kept that from working anyway. Accordingly, I commented that section out and ran another playtest. The only noticeable benefit is that by 1800 BC (and continuing for the next 300 years until I ended the test) turns were only taking a minute and a half, which is about one minute faster. I was also able to run the test with the "Debugslic=Yes" setting, so that confirmed the errors have been removed.

As a "control", I removed Frenzy from the next playtest and ran that down to 1500 BC as well. Turns are definitely faster - about 50 seconds to go through all 12 AI civs. As with previous playtests, the absence of frenzy led to more special attacks and better naval searching. The most obvious difference is that - aside from Nomads - civilian units are not included in military stacks.

City conquests were absent from both playtests, but there are contributing factors such as fewer continents, lack of a "true" human opponent (both tests used the "human-Barbarian" process), and the larger garrisons one typically sees in the improved Source Code version. I'll continue to look for possible AI improvements, but at least for now the game is solid and the AI appears to be competent in most areas. This revised Frenzy file will be packaged with the download, but will not be part of the default game - I'll leave that choice to each player. Anyway, it's definitely time to get back to the other improvement options that have been on hold for months now.

[Kull]

Even though most of my recent attention has been focused on the "big issues" (CTDs, Special Attacks, AI performance), quite a few "small" improvements have been implemented along the way. Some were simple, others pretty extensive:

1) AI Leader Gender & Personality: With few exceptions, until recent times the vast majority of world cultures have been male dominated, and their actual historical "leader lists" reflect that. For every Hatshepsut or Semiramis, there are hundreds of males. This is especially true with a predominantly "Ancient Era" mod like Cradle. Given that one has only two choices for Named Leaders to represent each culture from beginning to end, it's highly inaccurate to pretend that there's a 50-50 chance that one of them was - or could have been - female.

Accordingly, I've selected two of the more famous male leaders from each of Cradle's 33 civs and altered the gender sections in "civ_str.txt" and "ldl_str.txt" so that all pronouns and other descriptives are male.

Additionally, I altered many of the AI personalities as well, trying to match Leaders with their personalities (as far as we know them). For example, Carthage is now led either by Hannibal or his real-life rival Hanno. Historically, Hannibal was focused on aggressive conquest and is thus a "Strangelove" personality in CtP2. By contrast, Hanno represented the group which wanted Carthage to focus on farming and mercantile growth in the North African hinterlands, and is thus a "Medici". In general, most of the Cradle civs will have similar personality dichotomies, so the strategies that your AI opponents follow will in large part be dictated by their personalities, and those can shift radically depending on which leader is selected.

2) History Sections: There are now History entries for ALL Units, Buildings, and Wonders in the GL. Roughly half existed in "vanilla" CtP2, but Hex removed all of them. Those have been reinstated along with quite a few borrowed & edited from AOM, and a large number of entirely new ones created to fill in the gaps. It's mostly a "nice-to-have" feature but those descriptions which apply to the "future" are actually quite helpful as they provide the "back story" behind those otherwise unfamiliar units and structures.

3) Heptireme becomes Quinquereme: While researching the historical background for various units, it soon became obvious that the "Heptireme" (or Heptere) was an extremely rare design, achieved by adding a second set of rowers to the Quinquereme (or Pentere). Heptiremes were just one of many odd designs employed purely by the Greek Successor states (those which grew from the splintering of Alexander's vast empire), but they were always few in number.

By contrast, the Quinquereme was the true Iron Age successor to the Trireme. Thousands of them were built and employed by a wide range of different cultures, and they ruled the Mediterranean for centuries. Accordingly, Cradle 3+ will replace the "Heptireme" with the "Quinquereme". This is a "name only" change as all the graphics and statistics will remain the same.

4) Trireme Movement: The movement points for this unit are 3, where-as the earlier Bireme is 5 and the subsequent Quinquereme is 8. Even the Coracle has 5. By contrast (and as you would expect), the attack/defend numbers improve incrementally from Bireme>Trireme>Quinquereme. Having such a disparity in movement points makes the Trireme almost useless, and there's no historical reason behind it, especially when compared to the other two. Accordingly I doubled the range from 3 to 6.

5) Visible Wonders: Visually speaking, there are two fundamentally different types of Wonders - those with an "on-map" tile improvement, and those without. The TIMPs are quite valuable, adding significant food, production and commerce increases to the tile on which they appear (see attachment), and that remains even if the primary Wonder Benefits expire. Accordingly, it can be helpful to know which Wonders have a Visible component, so all their GL entries have been updated with that information.

6) Globesat Wonder: While adding "Visible Wonder" text, I noticed that the late game "Globesat" Wonder uses the Shakespeare "Globe Theatre" TIMP. Which is really strange, since there IS a "Shakespeare's Theatre" wonder already in-game...and it doesn't have a visible wonder! Anyway, I made the fixes to mapwonders.slc, gl_str.txt, and tilimp.txt so the Globe Theatre is properly linked and now appears with the building of Shakespeare (see attachment).

7) Lighthouse of Alexandria: Speaking of Wonders, the "Lighthouse" could be constructed anywhere, including inland cities. Which is an odd location for a structure that was designed to assist Naval navigation. Anyway, there is a building attribute used by the "Battlement" which restricts it to coastal locations, and testing showed that it works equally well with Wonders! So now - appropriately - the Lighthouse of Alexandria can only be built in coastal cities.

8) Actual Building Everywhere: Several of the Cradle Wonders provide the powerful benefit of creating the effect of a particular building in all cities owned by the Wonder-owner. Thus "Valley of the Kings" puts a "virtual Apothecary" in every one of your cities, the "Great Wall does the same with "Walls", and there's a few others which do likewise. This is controlled by the "BuildingEverywhere" attribute in the "Wonders.txt" file, but the downside is the AI can't process that information and will still construct the unnecessary building, even if it owns the Wonder.

As with many CtP2 bugs, a slic file was developed which prevents the AI from doing that, but there is a new attribute in AE which solves the whole problem. As the name suggests, "ActualBuildingEverywhere" creates the actual structure in every city owned by the civ which built the Wonder. It seems like a much better solution to the issue, plus it allows us to remove yet another slic file.

9) Garrison-to-Slave Ratio: In AE the AI is MUCH better at managing happiness in it's cities, and thus I haven't seen any spontaneous revolts in any playtests to this point. The AI is also better at using slavers, and thus you see a lot more cities with slaves. In addition, AE is coded to "spread the slaves around", so you don't have just one or two border cities acquiring huge slave populations. And lastly, the new AI is VERY focused on maintaining large city garrisons, which means that Slave & Unhappiness Revolts in AI cities are extremely unlikely anyway.

Accordingly, it seems appropriate to revisit the following settings in "const.txt", especially since (given most of the above) they are more likely to benefit the human player than the AI:
* UPRISING_CHANCE_PER_UNGUARDED_SLAVE 2 # from 5
* SLAVES_PER_MILITARY_UNIT 12 # from 3

Clearly those settings make it much easier to maintain large Slave populations in Cradle than the base game. My initial thought was to restore the AE settings, but instead opted for more of a "middle approach" and changed the "uprising chance" from 2 to 3 and the "slaves per unit" from 12 to 6.

10) "Illegal Move" sound: One of the interesting game features is that when the AI performs certain activities (not all), the associated sound is played, even when it occurs somewhere hidden by the "Fog of War". This is actually quite helpful, as it gives an audible clue to the human player that yes, the AI now has active Slavers or Prophets, that Piracy has begun, or that certain Diplomatic activities are underway.

The exception to all this is the "off-key piano chord" sound associated with an "Illegal Move". This was previously reported as an annoying issue that gets worse as the game proceeds, and I can confirm that it REALLY gets old. Not only is the sound rather jarring, but actually - who cares if the AI tries to perform an illegal move? Even when the human player does it, sound isn't necessary since there's plenty of visual indicators that, no, you can't perform "action X". Anyway, sounds.txt has been altered so that SOUND_ID_ILLEGAL_MOVE is now "NULL.WAV", and the "end turn process" is now a lot less annoying.