Making Cradle 3+ fully compatible with the Apolyton Edition

[Kull]

Game plays through to the end:

The timeframe for Cradle 5 (as with 3 and 4) is from 3000 BC to 2300 AD. And as you can see from the attached, it is possible to play the mod until the very end. Given the amount of time and effort invested in extending Cradle playability into the Future periods, that's a very important milestone.

For the past month, I've been giving Cradle a thorough playtest and at this point I'm confident that the mod is very stable (with a few caveats that will be elaborated upon shortly) and it's definitely a lot of fun - if you have fond memories of that "just one more turn" feeling, then Cradle 5 will be right up your alley.

​

[Kull]

Potential Issues:

There are some issues that could potentially affect your game, so let's go through them:

1) Breadbasket CTD: As I mentioned in the earlier post to Ekmek, this is a really nasty CTD, and the only workaround I found was to go into Edit Mode and use the Barbarians to conquer several of the AI civs before the game would finally proceed to the end. Fortunately this is a very late game CTD, and it never appeared in any playtest until the calendar exceeded 2100 AD. In addition, it occurred in a "robot" game (where the AI runs the "human" player) and because the AI is rather feeble when it comes to city conquests, everything about this type of campaign is "non-typical". So it's possible - even likely - that in a fully human-controlled game the player will so dominate the AI civs by this point that the conditions leading to the CTD will not be present.

2) Revolting Civs CTD: One of the more persistent late game CTDs was finally traced to a situation where an AI unit causes an enemy city to revolt and become a new civ (the "Incite Revolution" special attack), and then immediately deploys the same attack against the new single-city Civ. That causes the new Civ to revolt into yet another new Civ, but before the game has been able to process the existence of the first one. Four units have this ability (King's Eye, Spy, Secret Agent and Cyber Ninja), and the solution was to greatly reduce the likelihood of success while increasing the "death-chance" for each agent (percentages vary depending on the unit). Since that change was implemented I haven't seen any more of these CTDs, but it's still theoretically possible. The symptoms are that you'll see a message stating that "Civ X" has revolted and become "Civ Y", after which the game CTDs. If this happens in one of your games, one solution is to load an earlier save and use the Editor to delete the offending agent. Alternatively, edit the units.txt file to reduce the agent's chance of success to 0%, and then load the last save game. Run at least a few turns with the new settings and that should solve the problem (after which you can save, restore the original units.txt settings, and load the save).

3) Gigantic Maps: First of all, I absolutely LOVE playing with massive maps. The world just feels "bigger" and more realistic. The downside - again with "robot" control - is that the AI fills the map with units and cities and rarely eliminates any of the competition. So by the late game (1500 AD and beyond) the turns take forever and the game begins to experience the sort of instability (i.e. random CTDs) which is almost certainly game-memory driven. In other words, this is a game coded over 25 years ago, and it was never intended to handle a gigantic map (a size which was not part of the deliverable), much less all the additional TIMPs, Civs, and Slic code included with a mod of this size. In addition to the late game instability, the AI seems even more feeble than normal. However, most of this can be mitigated by selecting a "Regular" or even "Huge" sized map. Although once again, this may just be a side effect caused by using "robot-mode" for the human player.

4) Pace of Technological Change: This is affected by the number of civs and even the map size you choose. I've never played "one-turn-at-a-time" beyond 1500 BC, but in "robot-human" games, the technology proceeded at close to the historical rate with 12 civs on a Gigantic map. Conversely, 15 civs on a Regular map reached only the Modern Era by the time the game ended in 2300 AD. Just be aware that your starting choices will have ramifications in this area.

5) Rare "Upgrade Elite Unit" Bug: This has only happened once in 100's of playtests, and the requirements are rare enough that you'll probably never experience it, but here it is: If you have a single "Elite Spearman" unit (there is no bug with 2 or more) AND he is not located in a City or Fort when you discover the "Bronze Working" Advance (which triggers the Upgrade), the Slic code will deliver some bogus messages and will NOT move on to the "Militia Upgrade" options which should follow immediately thereafter. If this happens to you, the best option is to load an earlier save and move the unit into a city or fort. Failing that, disband it. The optimal solution would be to re-write the code in updater3.slc, but I can't spot the problem so unfortunately it's not an option.

6) "Lost Tribe" related CTD: There are several possible results from "popping" a Goody Hut, such as gaining Gold or a new Unit, and the range of options and associated percentages are controlled by the risks.txt file. In the standard game, one of these options is to find a "Lost Tribe" and thereby gain a free city. However, in very rare instances the AI will build a city right next to one of these "Lost Tribe" Goody Huts, and popping THAT causes an incompatibility with the game code (which prohibits founding cities in adjacent tiles). That issue - possibly exacerbated by Slic Code - soon causes the game to CTD when being saved, so it's a serious bug. Accordingly (after a brief flirtation), Cradle 5 now sets this chance at 0%, so it's not a CTD you have to worry about. However, I know people like to tweak their gameplay settings, and the "Lost Tribe" Goody Hut is a fan favorite. But in this case it is too dangerous, so don't make it part of your Cradle games.​

7) Visible Wonders "Bugs": Sometimes the player (usually the AI) completes a Visible Wonder, but the graphic does not appear on the map. Although infrequent, this can be caused by several things:

- No available terrain to place it on. In the most extreme example, a single hex island has no adjacent land hexes, so Visible Wonders will never appear if built by a city located there. More common is lack of suitable terrain (for example, only mountain or glacier terrain). Another possibility is that all the suitable terrain is already occupied by other Visible Wonders ("normal" TIMPs will be overwritten if nothing else is available).

- Although this next occurrence is rare, in some instances the AI completes the Wonder and has suitable open terrain (unoccupied by any TIMPs) on which the code can place the Visible TIMP. But because the AI is an idiot, on the same turn that it completes the Wonder, it will start building a DIFFERENT TIMP on the same hex. Even though it is impossible to start building a TIMP on top of a COMPLETED Visible Wonder, it is possible to do so in the single turn it takes for the Visible Wonder TIMP to be completed. And when that happens, it will replace the Wonder TIMP.​

[Kull]

Final set of Improvements:

The extended playtests (both robot and fully human) identified a few improvement opportunities:

1) Visible Wonder for Ramayana: This wonder has a simple effect (+1 Happiness until Feudalism) but now includes an "on-map" component. Even though it's effectively the "Book Wonder" for the Hindu religion, it doesn't have as many benefits as the others (Islam, Christianity, etc). In this case we'll use a Ziggurat TIMP from the Visible Wonders mod, which Hex modified to add trees and other structures, thus giving it an "Indus Culture" appearance appropriate to the Ramayana Era. Files changed: tileimp.txt, mapwonders.slc, Great_Library.txt, gl_str.txt & gtset555.til​

2) New Great Library: My attempt at a "Great Library" Visible Wonder graphic, while perhaps technically accurate (nobody knows for sure) was too fuzzy and washed out compared to the others. It has now been replaced with a new one based on the "Baalbek" graphic from one of the Civ3 modders.

3) New Taj Mahal: The previous graphic came from the old CtP2 Visible Wonders mod, but it was small and the image quality poor. This too has been replaced with a new one (edited for size) from one of the Civ3 mods.

4) The "Colony" TIMP (an Egyptian-style pylon surrounded by houses) looks very dated by the late game. Once again a graphic was borrowed from Civ3 and the numeric link in tileimp.txt will be switched as part of the "Modern Era" file swap.

5) City compatibility with City Wall: Cradle 5 has a lot of new cities, and most of them look really nice. But what I hadn't considered was what happens when the player constructs the "City Wall" building. This creates an on-map graphic - a smooth stone wall surrounding the city - and that provides a visual cue, warning the player to expect improved city defense against both military and slaving attacks. The problem in this case was that 3 of the 4 ancient-era Celtic city sprites were off center, and thus appeared to "float" over the city wall graphic. And the same was true with two of the Mycenaean cities. All are now fixed.

6) New Mesopotamian cities: To be perfectly frank, the 4 city sprites I created for the Mesopotamian Medieval & Renaissance eras were just awful. Instead they are being replaced with 4 new ones taken and edited from a Civ3 city mod (see attached). In addition, the Ancient Era will now begin with a central temple (the ziggurat comes later) and the original levels 1 & 2 will become 2 & 3. The Classical era will continue to use the original 1, 2, and 3 (so no change). This does require an additional city sprite slot (since we're at the maximum), once again taken from the "Palace" group. File changed: agecitystyle.txt