AdvCiv-SAS (Simple Advanced Strategy)

[civ4-advciv-oracle-bug]

Hello

Good to hear that you're enjoying it i mean if i may say.

Interestingly, it seems i am going in the quite opposite direction xd if i may say too, as i have implemented more or less most of what i wanted in advciv-sas as of now (except maybe the future era music, or some feature and tweaks that i'd need to dig into, but otherwise i believe the ai is smart enough and really overall competent xd really but anyways etc in my mod now )

So i was actually starting to start some new hobby instead xd; as for the ressources that may help you gathering bits from my code, consider the following (maybe slightly outdated but hopefully not too much):

- known issues: https://github.com/wonderingabout/A...es/README_Known_Issues_In_Base_AdvCiv_Civ4.md
- very quick start guide (chatgpt written, more readable perhaps): https://github.com/wonderingabout/A...ME_Very_Quick_Get_Started_Guide_By_ChatGPT.md
- quick guide (maybe refer to this for extra details of same info more or less, but more subjective and less readable xd): https://github.com/wonderingabout/A..._Appendixes/README_Quick_Get_Started_Guide.md

One of my last tasks may be to add some defines, so that say if players want a base advciv feel, they can disable some of the code chunks i added (the relevant ones), like smart buildings, smart military, no production fallback, although i think mine are vastly superior xd, in some cases users may want to default to base advciv, so i'd need to add a toggle for these in defines ideally (but tedious so not sure how far i'd go doing it but hoepfully quite a bit).

As for what may help you, here are the ideas that i think you may want to implement from my mod (i guess stability and bugfixes or conservative tweaks first right? (may be updated later):

- workboat loop fix (very nasty and affecting all civs including i mean not only barbarians but anyways etc)
- no citizen specialist
- possibly also no early specialist tweaks or such as ai is very inefficient and does very bad things with them xd
- settler national unit
- large first city settling window (5 free turns to choose best city i think more or less if i remember it correctly)
- settler founding value and such helper functions as well logic (i removed or rewrote some), read code comments if intelligible and you are patient enough xd hehe thanks althgough may not be too easy but hppefully not too painful or tedious xd but as you prefer i mean as can be hard i mean i guess if i may say
- worker logic (i guess our workers are just much much much smarter and efficient, although sometimes they don't road enough, overall they do very well i guess and just are so much better than base advciv +/- civ4 ones i think you may want to implement these as well (as you prefer for this as well as i changed many places but hopefully not too hard to find xd) (nextcitytoimprove, workermove, bestcitybuild, and related functions i guess)
- i guess the sevopedia reworks xd maybe too if i may say but anyways etc, as my sevopedia is super pretty and informative (if not messy sometimes but anyways etc xd anyways etc), might need minimal DLL changes but otherwise mostly python and compatible with more or less any civ4 mod (they can add their mod specific things like new xml fields or such i guess)

Probably many i am forgetting but hopefully these are a quite good start xd but anyways etc

If you're curious you may want to look into the scripts as well, as i believe they are quite easy to set up (see related documentation or main readme), and showing handicap table for example can be quite nice i think (if you changed some fields i guess you can minimally update the script to support them as well if i am guessing correctly that my script would support these as well with minimal updates xd)

But overall also thanks a lot for your feedback, although i knew my mod achieves quite a lot xd i mean if i may say but anyways etc, it's always good to know (i guess, maybe, in this case i mean but anyways etc) that my mod is useful and fun or something like this or a combination of these or such xd thanks but anyways etc thanks

Also as for integrated ais like cursor or such, i actually never used these, could be fun and interesting xd, although i'd be a bit cautious about ai going all over the place with uneeded extra info unrelated to solving the immediate issue i guess maybe, but i could be mistaken as i never used it as such, just some people mention this and it seems to align with my intuition too i mean if i may say. And also because i more or less finished developping advciv-sas, although it could always be improved, and i could (or not but anyways etc) add some tweaks or defines, but that is not too sure xd, i may or may not but anyways etc

 

[Rawwwrr]

How difficult will it be to apply your AI modifications to a regular AdvCiv?

 

[keldath]

Rawwwrr​

its not hard, all you need is winmerge.

but , sass is on top of advciv already.
do you mean to some modified version of advciv?

 

[keldath]

civ4-advciv-oracle-bug​

hey,
"Also as for integrated ais like cursor or such, i actually enver used these"
in my work i have Cursor and the agents, mostly Claude, which is magnificent, is working with the repo.
and with contexts of the designed changes it can do and understand everything and how the functions flow

.i haven't yet linked my repo with gemini 2.5 flash (free), but i might do so.
the results are far greater compared to pasting code to the web page of the ai's.

--
anyhow,
i went over plenty of the dll code you did.
some cool stuff. very specific case bulks which i assume as you say saw issues and asked for a correction.
i merged some stuff into my mod.
what i have not , are parts where you have hard coded stuff, like the improvements and such.
having hard coded stuff is not recommended , especially in mods, but i see why,
yet, im not sure this was needed with you code. i suggest just have the code loop over each improvement and create an enum of improvement and the bonuses it is linked to.
i would imagine this: {mine: [iron, copper, sulfur....etc].
then apply the code for the ai from it as you did.

i looked at the settler and workboat , i need to lok at it more to see what ill take from it.

looks like you did tremendous job, i salute you and hope to see you create more.
if ill see some more stuff, ill let you know.

 

[Rawwwrr]

I was referring to the regular version of AdvCiv. I won't be able to upgrade myself, my skills are too poor for that.

 

[civ4-advciv-oracle-bug]

Yes if you're using advciv 1.12 (note: the latest stable is base advciv 1.11 on thread you linked, consider using latest base advciv 1.12 for most fixes and tweaks since then, ssee below at my edit comments for details anyways etc) already it shouldn't be too hard in itself just it might take some time (quite some time actually if you want to do it carefully and exhaustively, looking at other related changes or such, but in itself not "hard", just potentially quite lengthy i mean if you want to implement many of my changes but anyways etc).

If you know how to compile the dll, it's all about handpicking whichever changes you want from my mod AdvCiv-SAS into base AdvCiv 1.12, since AdvCiv-SAS is based on as of now latest base AdvCiv 1.12, as i assume @keldath is doing or considering to do. Just i would use VS Code for that rather as you have a compare files feature that is handy, but there may be even better ways than that that i don't know about as i don't know that much about these and mostly learned on the fly or from some people who taught me thanks btw xd really but anyways etc.

Also, if i find the motivation to do so ideally, i'd add more and more defines, so that you can disable some features i added if you prefer, as i started doing here: https://github.com/wonderingabout/A...ework/Assets/XML/GlobalDefines_advciv_sas.xml.

edit 3: The advantage of using xml defines rather than hardcoded values in the DLL source code, when possible, is that you can change the values in the XML (when it makes sense in the code to do so, e.g. one value that stays the same for all games, you can extract it out of the DLL code and say "use the value in the .xml defines file rather", then it can be changed without a dll recompile nicely then, but which wouldn't make sense for values that depend on game state or such as chatgpt 5 explained to me xd i mean although i coudl guess so but maybe helpful to state clearly here and easy to mix up so i like to ask chatgpt 5 to review my changes or other ais if needed but anyways etc) (you'd need to restart the game to apply changes though if i'm not mistaken, and make sure to not use nonsensical values like -99999999 or "abcde" which could cause weird results xd unless supproted or relevant, but otherwise should be fine xd (but test and check to be sure and i am not responsible altohugh should be safe but check to be sure or ask some ai or such if you'd want to use some unexpected/weird values but anyways etc)

But most of my changes are pure enhancements, so i don't see much point going back to base advciv, except maybe using the old xml stats of units or tech tree, otherwise it's mostly just fixes or enhancements that i tested ingame to be better for most if not all: you don't want 3 cities in 10 to be stuck trying to produce a workboat then scrapping it from 10 to 50 turns even for non barbarians, nor cities stuck at 2 pop running 1-2 scientist(s) or a settler for 50 turns in 2 cities, etc etc etc, i see so little upside to revert to base advciv that i don't think i would even add a define for these.

Ideally i'd add some more defines, but i am pretty much done for advciv-sas it seems as ai is much much better/stronger, the game more balanced, really challenging with less handicap than base advciv overall but much harder challenge, all as i wanted xd. It could always be improved, but looking into new hobbies or such now it seems At least in theory if i may say but anyways etc.

For defines, you may want to implement your own defines, if so, you'd need to add your file here: https://github.com/wonderingabout/A...GameCoreDLL/CvXMLLoadUtilitySet.cpp#L197-L200

Ideally i'd add the citizen specialist no auto governor assign as i think it's very annoying, even more ideally per specialist no auto setup (like no auto spy, no auto artist, but ok to add auto scientist or auto priest by governor (manual controls would remain only the auto thing), it's not hard and i already did so for the AI for the citizen=default specialist and some extra optimizations, just need ot motivate myself and use the time for it xd. So i may or may not do it, not guaranteed, but anyways etc. Still, i hope this info is helpful but anyways etc.

Of course i don't know if i would develop AdvCiv-SAS more as i said that before and added more thing,s but i am satisfied so far of what we did, and we don't have that much players, still it's fun as it is and really good, although could always be imrproved. AdvCiv-SAS is strong, and likely the strongest by far of all civ4 mods AIs, although i didn't test them all, it's very smart at least much more than abse advciv 1.12, othe rmods may be interested in importing some bits from it fine with me hehe as said in github readme doc but anyways etc (refer to there for details on crediting me and ais who helped me if you want ideally (not obligated) but anyways etc).

Edit: consider switching to advciv 1.12 rather: https://github.com/f1rpo/AdvCiv/tree/1.12 (download zip from github website (click on rectangle button then download zip)), it should have many more fixes like the oracle bug xd of when i started this civfanatics account xd and likely many others as well but is just my opinion and i didn't look in detail at which changes were made, just it seems stable enough to start from there rather if you want to do your own modifications.

Edit 2: the current latest stable advciv you linked is advciv 1.11 which is old and lacks many changes as shown here: https://forums.civfanatics.com/threads/advanced-civ.614217/post-16758705 by fellow @f1rpo xd before he/she(/it? Xd who knows but anyways etc thanks for all regardless if i may say in this case i mean but anyways etc) left but thanks for all i mean really but anyways etc; there are many changes so it's worth using latest version rather i think but check to be sure anyways etc.

 

[civ4-advciv-oracle-bug]

@keldath thanks for your feedback as well

Yes ideally i would have done it more cleanly, but i don't know that much C++ actually, plus i exported or started to export some values in sas defines now for example the aifoundvalue function should be fully customizable which is nice: https://github.com/wonderingabout/A...ework/Assets/XML/GlobalDefines_advciv_sas.xml

As for terrains and features, the scope i had in mind for my mod is small, and it would very much complicate logic to, say, if i understand it correctly, "among all improvements if it gives 1 food or more", or say like "among all terrains or features, if they have at least 2 natural + improved food" etc., it can quickly get very complicated, and we don't change them that much if at all in advciv-sas, so it is much easier to do a per terrain/feature valuing (e.g. "for grassland if hill do that / value it as that, else do that / value it as that; for plains if hill etc."), however i could have exported the values in xml defines rather for external tuning.

If a modder wants to reuse my code, for example to add new terrains, i believe it's not too hard to generalize my logic, or simply add their own custom logic for their new terrain. Ideally i would have exported values for workers for example in xml defines though, but otherwise i'd say i believe it is also defensible to do as i did, although i don't know too much C++ to tell seems fine as such i'd say But anyways etc.

For buildings however for example, they are not that hard to describe as such, and can change quite often, so code is almost entirely blind (except things like palace or such) so that even if players change buildings, ai would auto adjust based on latest xml (e.g. "the building(s) that give 1+ happiness" rather than monument classes or colosseum classes buildings that are not hardcoded unless badly needed which i don't think i did but anyways etc, etc.).

Again i'd say to be bit cautious about ais, they are great to help code, especially since i assume you know much more than me in C++ or such, but if you're using the free version, it gets much less free time, depth of thinking, context window, etc. I am not sure you can trust it too much or get best outputs. Even in free mode, i believe you have access to a few gemini pro prompts per day, same for other ais generally, consider using these rather while carefully stating your purpose in the prompt depending on what you do. I have found chatgpt 5 and grok ai to be most reliable for my needs in their expert or deep or extended thinking whatever they call it modes rather.

In free/fast it would spit the first thing it can think of, without pondering on the problem too deeply, so i'd again advise to be very cautious, even if you know what you're doing, you'd have higher quality code and thinking/logic and solutions tailored to your problem if i may say if you can use a more advanced version. I guess agentic AI must be nice, but i also read it's the same technology bundled in another UI than using catgpt website for example, some reddit users seem to say on a quick glance that too much context is actually not good, but again i don't know enough to tell, however my observations while using said and other ais seem to line up with this: the less it knows the better, just give it enough relevant samples so it knows code specific style and patterns well enough (like 2003 code, some style of classes or such which i don't know too much about, it can read code comments too so i hope mine help as well if you feed them to it) (maybe documentation helps ai too if updated enough but anyways etc)

edit: the workboat loop issue is problematic, it affects all AI (not the human player if i remember it correctly) players not just barbarians, and is not that rare, see screenshots linked in known issue for diff if i'm not mistaken but anyways etc. But you may want to not just implement the no scrap local workboat function fix (from what i remember of it), i believe i changed other things related to this, but i'd say documentation is your best bet as i most likely talked about this (or in code comments); see known issues documentation as well for extra info if i added it there as well.

More generally sure, feel free to ask any question, if i can help i'd be glad to (as long as i read messages here ideally i mean but anyways etc, civ4 modding is fun in itself i'd say although i'm mostly done (i guess as well but anyways etc))

 

[onkelm]

Hi, looks all awesome !


A question for modding: is it hard to add some features ?
- use mountains for special units
- advanced trade,like buy units, sell units, etc.
- surround and destroy mechanics
- ranged attack

because this mod could be the perfect base, to devlop a ww2 mod from that

I would start such a project, but I do have very limited c++ experience.

 

[civ4-advciv-oracle-bug]

Hello

Thanks for feedback/questions

General answer is i don't think it would be too hard, in fact some mods already have features like these although i can't name them in detail (look, from a quick search and what i remembered of it (although i have not tested them to know for sure) at mods like https://forums.civfanatics.com/threads/advanced-diplomacy-2.447915/ or such if they help maybe and see how easily it could fit in your mod), but as you explained as well for yourself, i also have limited C++ knowledge xd. Still, AI may provide some help if you have some basics or experience with another programming language possibly.

Among the features you listed, the easiest one seems to be to use peak/mountains for units. I added in sevopedia a peak (and hill too) item so you can see which units have access to peak (see screenshots appended in this post as well).

For example, by modifying relevant xml fields like bCanMoveImpassable such as here https://github.com/wonderingabout/A...ssets/XML/Units/CIV4UnitInfos.xml#L7431-L7432 for example now these units are able to walk on peak.

In AdvCiv-SAS, i allowed some units as shown in as of now screenshot below to do but anyways etc, as i found the mechanic interesting. I had also considered to allow city settling there at engineering like some mexican or south american or other cities in the world are (at high altitude if i'm not mistaken), but it seemed bit tedious and i was not sure it was worth the complexity addition as former @f1rpo at least as of now but anyways etc maintainer explains in some code comments or such xd thanks, and i felt the same about these hehe if i may say but anyways etc, so i held off and didn't do these, but they are interesting ideas i think.

I didn't test it, so i don't know if you could say just settle on a peak tile if a settler can walk there, but if you test it i'd like to know xd if i still read here and you test it xd

You can already do some changes just with such XML fields, if you encounter issues or want more advanced changes, it may indeed need/require some C++ or such DLL coding rather (you can export to global defines some of your constant variables like options to enable/disable, specific values so you can tune them in XML without recompile required, etc.).

As for surround and destroy, i don't know what it means, but if you're referring to zone of control, i don't think (but i may be mistaken, just a guess / vague rememberance anyways etc) civ4 has it, and it seems fine as such, also i don't know how to add it xd. If it's something else, and as for many of your suggestions, consider using AIs cautiously to help you spot where to make targeted conservative, safe, minimalist changes that help you gradually or in one go implement what you want, possibly by taking some inspiration from what others mods may do or not as i linked above (but untested, check to be sure)

For ranged attack in particular, i also wanted it at some point (adding a sniper unit for example that would attack from a distance or such), but i couldn't find easily how to do it (it was before i knew or started to modify the DLL, but i assume it is maybe possible), still, as i developped AdvCiv-SAS based on advciv 1.12 but anyways etc, i have found less units is better for my needs. I had developped a quite lengthy and ambitious if i may sya but anyways etc military tree rework with many new units (https://github.com/wonderingabout/A...about-a-more-advanced-prototype-military-tree), until i realized a simple tree with minimal units is much better for my needs and this mod hehe and suits me better mod doesn't need to be bloated to be hard or fun i think, and i liked the simple minimalist idea of it But anyways etc xd anyways etc.

If you specifically want to develop a ww2 mod, i can see (i think/guess?) where the enthusiasm may be, as AIs in AdvCiv-SAS are super strong and pump many units efficiently (at least quite/very efficiently), with forgiving/lenient handicap settings, but very brutal and strong AI so it doesn't feel a grind but still very and much more challenging than base AdvCiv at same difficulty xd, so if you were to do a ww2 like mod, it may really feel like an epic world battle, however i would just add some caution about this is that you'd need significant tech and unit and buildings rework, the scope of such a project is very large, unless you delete only units and builidings, plus some code changes i made rely on era, and some parts of the code do as well i think too (but check to be sure), so you may want to either keep era modern, or subdivide in eras, etc, i don't know the proper way to do it, as if you start with pop 1 cities you'd still want early eras with growth priorities i guess.

But if you only want to keep more or less my tree but add your own extra long modern techs, then it might not be so hard or require extra changes, but you may have to maybe tweak tech costs like making modern techs much cheaper but with much more techs, it might work this way with minimal reworks possibly and no need to handle the early game where you have few cities, but it's just a suggestion.

Still, it's gonna / would take quite a lot of time, so consider this before diving deep into it, think about what you want to do, the scope of your project, why maybe you'd want to do it as it may maybe guide how you'd want to implement changes or which changes (but mine changed over time without clear early guideline except from the desire to fix and enhance things to make them simpler, fairer, and harder mostly xd so i mean do as you see fit/prefer but anyways etc)

Good luck hehe, but consider/ponder where you'd want to go with this, also civ4 is not as popular as before it seems at least as of now, but it was very fun developping advciv-sas, and although as of now i don't play it anymore if i may say but anyways etc, even autoplay is quite impressive to watch (if i ever do, which spoiler i don't do so much if at all now xd since i'm not looking to improve the mod further as of now it seems (who knows about the future, but it is significant time investment, otherwise i have some ideas, but tedious to implement (i even have some future era music ready to add!)).

If you're doing it for yourself whatever that means it may be a fun experience possibly, but if you'd want a mod to be super popular, it really seems like civ4 has less players nowadays / quite recently, but it's not like there are none, and i really enjoyed developping advciv-sas, so don't mind my advice/feedback too much on this, but i'd say to maybe consider/ponder over it and do as you see fit hopefully this helped a bit thanks for your question and such anyways etc thanks anyways etc

 

[civ4-advciv-oracle-bug]

also i have tested quick enabling this moveimpassable xml field for settlers, and even if they can walk on peak, they still can't found a city. Super fun to try xd. I assume either in XML or in c++ you may or maybe shoudl (but check to be sure as it's just a guess of mine if i may say but anyways etc) be able to found a city on peak (maybe, again check to be sure xd but anyways etc)

 

[civ4-advciv-oracle-bug]

Re on this, i actually added (not committed yet) the ability to settle on peak, it was actually super easy in bool CvPlot::canEverFound function, and added a define so it can be optionally disabled. I'm not sure if default should be to enable or disable it. It's quite altering so may go for disabled, but it's not that game changing actually (may have to choose a non op defense modifier for peak at best, city tile yields are as of now low so i may leave it as such, and tell ai to slightly value it as home plot if the define that settlers can found there is enabled (and if they can walk there too on the coding end but anyways etc))

Result as in screenshots below xd, thanks for feedback, i wanted to add it long ago, but i had no idea it would be so easy if i may say but anyways etc.

I was supposed to not develop this mod anymore at least not yet (bit tedious xd), but this is a small addition so why not in this case i mean if i may say but anyways etc.

edit 2: not sure i'd add it, is bit tedious xd although not too long to add, will see xd but anyways etc.

edit 3: i forgot to account for the fact that if settlers can found on peak (or ice cap since it's impassable too if i'm not mistaken but anyways etc), then all units have to be able to walk on peak (or ice cap etc anyways etc) else the new added feature/option is useless xd as cities can't be captured, so i may scrap the whole idea actually or implement it with that caution in mind for players but anyways etc

edit 4: it's also tedious because i'd need in AIFoundValue::evaluate to incorporate peak calculation into home plot valuing, which if i do too soon in the plot loop or too late, may indirectly affect how it is computed (i don't know for sure, but i suspect impassable plots are rejected since advciv or civ4 didn't expect settlers to settle there (but is just a guess so may be mistaken, check to be sure anyways etc), but this may create unexpected issues if impassable are computed in functions or such, so to be safe i may do it indirectly (maybe, again just a guess and to be safe, may or not be needed or accurate but may be, check to be sure anyways etc) rather in one way or another at the early part of the loop i guess (but is just a vague guess) but anyways etc, still i'm not even playing it xd so not sure it would be a significant addition, players would most importantly have to enable moveimpassable to all units else cities would be unattackable too as well. It is maybe a good example of too broaad scope change xd that does not significantly enough change gameplay or meaningfully enough to justify altering existing gameplay which is fine as such (peak giving strategic positioning as well not only bad yields, plus some units can walk there as of now in advciv-sas, vs no such advantage if all can, so all in all it seems better to scrap this feature xd but anyways etc, however i'll add the code in another git branch for reference or such maybe i guess but anyways etc).

 

[keldath]

hey folks,

well i added in my advciv mod:

A question for modding: is it hard to add some features ?
- use mountains for special units
- advanced trade,like buy units, sell units, etc.
- surround and destroy mechanics
- ranged attack

i have mountains passable
trade - i added also a new trade, its doable to make a unit trade, but not easy.
surround and destroy -> i have unit blockade -> which is a zone of control
ranged -> very complicated -> i have tried several versions -> some worked, like in realism invictus , but the ai is meh. so i implemented a mechanics that empowers ranged units.

civ4-advciv-oracle-bug​

you can easily merge from my mod to yours stuff from the above or more.

 

[civ4-advciv-oracle-bug]

@keldath thanks but i think i got it mostly figured out

Details in: https://github.com/wonderingabout/A...161c1f200535b186ebc21b488c2252b46b?diff=split

I don't think i'll implement this actually hehe, it has many restrictions for players, and questionable gameplay gains, although it may be cool to settle on peak, we'd need to handle ice cap which is less realistic, if i'm not mistaken but anyways etc, plus, peak or such add some strategic value for military combat or such or settling preferences, so maybe keeping it as such is easier and fine, still i added the code quick for fun but it doesn't look wise to invest further on this issue

I may look into your mod if i need some changes (but again i like to do my own things generally xd in very simple terms as i can understand or/and enjoy them but anyways etc), but for this issue it seems fine as such, plus i'm not developping advciv-sas further it seems at least as of now xd if not always or not or yes or etc but anyways etc

Still, i hope this code feedback is helpful (else i still did it for fun i guess) but anyways etc

(note: untested on ice cap but i assume/suspect it woudl hopefully work same xd but anyways etc)

(note 2: as of now still partially unfinished for the very bad tile count fix (if we settle on a very bad tile like peak or ice cap, no need to count it as such at city site evaluation, but i didn't implement it yet as tedious, otherwise should work as is minus this slight but still ideally to fix but anyways etc miscalculation if i'm not mistaken i mean (also not reviewed by ai i just coded it myself in one go hehe with bit testing and extra logic gradually added and tweaks too later as i went but anyways etc)

(note 3: again (very) unlikely i'd implement it, it has too many drawbacks for little gain i guess/think after evaluating it in this case i mean but anyways etc), but added code for reference but anyways etc.).

edit: about the other features, and thanks for feedback about them too i mean but anyways etc, worker trading may be cool like they do in advanced diplomacy 2 or such, but it may alter too much gameplay, i'm fine with simple diplomacy for my mod (and most importantly not developping it anymore it seems as ai is super strong enough and i want to focus on other hobbies xd or such, although i don't mind reading here everywhile or such as well i mean if i may say at least as of now but anyways etc), i think focus should be (at least for my mod but anyways etc) on actualy strategy, game is otherwise fine as such for scope of my mod, although some features may otherwise be cool, they are not my focus, i'd rather simply want AI to be as strong and efficient as possible, with otherwise simple gameplay mechanics unless i find them relevant to add depth/strategy/challenge in a meaningful for my needs/preferences/vision whatever it means etc but anyways etc needs i mean if i may say (but not developping advciv-sas as well i mean too if i may say it seems but anyways etc) but anyways etc.

edit 2: it's actually even more tedious than i thought, because then i'd need to also consider worker builds (can we build roads here, at which speed, is it tunable, etc), combat modifiers (not op but not weak), which units can walk on it (can catapults or siege or mounted units climb rocky terrain, etc), etc. My code was also more wrong than i thought i mean but anyways etc as ice cap does not apply to land units so no point to settle a city here or it would never be captured (unless by galleons or galleys or whatever which is fun but like a super mod mod xd at this point i'd say (and so i'd rather focus on core/main focus which would be as of now only impassable peak, not impassable ice cap, but even that is tedious i mean but anyways etc), see as of now below in this edit 2 part of the post for details but anyways etc) but anyways etc). All in all, doesn't seem so worth to add due to complexity xd, and the core gameplay should be in fighting other civs so i'd like players to focus on that rather, however this helped me fix a wrong assumption in sevopedia ice cap, where scouts or such were shown to be able to walk on ice cap which is false since it's a water terrain (so a domain land unit would need canmoveallterrain not just canmoveimpassable, but a domain sea unit (untested but i assume so based on chatgpt 5's feedback as well but anyways etc) submarine could walk on ice cap just with canmoveimpassable without needing canmoveallterrain (and it would still not be able to walk on land with just canmoveimpassable if i'm not mistaken (but just a guess, untested ingame anyways etc)) so i added this fix in the main mod branch. I'm reluctant to involve myself further in the mod xd, doing that was quite tedious, and although bit fun, i'd like to focus or do other things i mean but anyways etc. It's too bad because it's really fun (but also painful too at times xd but anyways etc) but tedious so maybe is just how it is or not or yes or etc but anyways etc, in all cases i added the fix here for sevopedia feature ice (cap) but anyways etc.

 

[civ4-advciv-oracle-bug]

Hello, this is strange, i tried it and it seems to work as intended (julius caesar being aggressive and imperialistic) as shown in screenshots.

Did you clean delete the entire advciv-sas folder (if you had a previous version)?

Or maybe you perhaps downloaded another version than 5055 without noticing?

I'd suggest clean delete the entire AdvCiv-SAS folder at (for example in my computer using Steam, adjust if your civ4 Program Files or such folder is elsewhere) the folder to entirely delete each time you want to change AdvCiv-SAS version is for example at: C:\Program Files (x86)\Steam\steamapps\common\Sid Meier's Civilization IV Beyond the Sword\Beyond the Sword\Mods\AdvCiv-SAS\ (save files are in the windows user account folder so should be unaffected)

If it fails and still same issue, consider redownloading AdvCiv-SAS 5055 (or whichever version is latest)

hopefully helps anyways etc, if not tell me np if i can help xd thanks

Edit: wait, you're talking about praetorians, but in AdvCiv-SAS (and AdvCiv) rome empire's unit is the legionary not the praetorian, maybe you're not running the mod itself?

Windows shortcut should be like this (for example for Steam, adjust if your game path is elsewhere): "C:\Program Files (x86)\Steam\steamapps\common\Sid Meier's Civilization IV Beyond the Sword\Beyond the Sword\Civ4BeyondSword.exe" mod=\AdvCiv-SAS (windows steam shortcut also included in AdvCiv-SAS mod folder if helps you can just copy paste it as you prefer anyways etc)

note: unrelated to your current issue as AdvCiv-SAS 5055 should have the new trait, but alternatively, you may want to try the development version if it has some new fixes or such (in case you'd want latest features (there are not so much since we are only at AdvCiv-SAS 5062 (and some changes are mostly if not only documentation without gameplay change but there are as of now a few gameplay changes as well since AdCiv-SAS 5055 but anyways etc) as of now, but may as well try it if you want too as an option i mean): https://github.com/wonderingabout/A...ck_Install_Setup_Guide.md#development-version

edit 2: there was a comment here which i replied to if i'm not mistaken, but it disappeared somehow, well hopefully this answer helps as general guidelines or/and fixes in similar or install issues or such or whatever xd or not or etc but anyways etc

 

[civ4-advciv-oracle-bug]

Btw i'm thinking to rework sevopedia leader a bit to add a 4th column, so we'd have more information to display (currently some items are hidden due to space constraint, others block back and next buttons which is not practical too), and which may be quite nice.

I've come up with this draft for the new design (idea xd but anyways etc) quick xd but anyways etc, i find it cool, only downside is description would need to be moved to another place, but i think it's a fine trade off, as it's least useful info, once known i don't think players reread it many times, and even if they do, they could go to an encyclopedia like pedia page per leader item in sevopedia.

I keep saying i want to focus on other things/hobbies (which is true), but at the same time i think it would be a quite cool thing to add too hehe (not promised though, i may or may not do so but anyways etc), if i'd maybe (not guaranteed as well i may or may not do so but anyways etc) add a few extra features, then i may make a release to have all these hehe.

If you have feedback welcome although i can't/don't guarantee i'll listen to it as i have my idea in mind xd if i may say but feel free to share or not xd but anyways etc

edit: (an)other downside is it's tedious though to move all leader bio entires to another page but may be worth it for extra column and readability + navigation back/next quick between leaders which may be suepr useful ingame by for example selecting a few leaders we want to comapre quick or such (even from main menu actually but anyways etc), but still bit/quite tedious xd, i cant promise i'll/'d do it, but i thought i'd sahre the ida because why not but i may or may not do it but feeel free to tell your opinion or not but anyways etc

 

[civ4-advciv-oracle-bug]

new version with much stronger AI

AdvCiv-SAS 5082:
- Much stronger AI than in AdvCiv-SAS 5055, in particular it should be way more resistant to snowballing and better at defending and going into the late game, leading games to generally be closer and harder to decide on a winner early on:

• improved worker AI efficiency (apply 1 worker per tile logic to more functions, attempt to improve cities in more cases rather than idling) and reduced parking in cities,
• increase siege threshold pre renaissance if we have no key strategic bonus (as of now iron, copper, horse, camel, elephants) (we need more catapults and trebuchets if we have no key strategic bonus, helps reduce longbow spam, and make a successful longbow + catapult +/- trebuchets rush)
• force one artist if we don't have BFC and city culture per turn is low, until we do, in CvCity::doTurn
• force emergency buildingclass_harbor if city is coastal (ocean) and low food per turn (tunable in sas defines (https://github.com/wonderingabout/A...ework/Assets/XML/GlobalDefines_advciv_sas.xml) anyways etc)
• add emergency defense buildings to build at highest priority if invasion risk is high enough, in CvCity::doTurn: This seems to lead in autoplay to much harder snowball, and to games lasting significantly longer. As a side effect, it also reduces a bit unit spam which should slightly improve economy and bankruptcy risk. This change should make AI players especially stronger, at least it appears so in autoplay.
• make civ specific unit importance evaluation lower and tunable, and other tweaks or fixes: addresses the case of japan ai relying too much on samurais when knight units are available, and now it is tunable in sas defines.
• enhance AI promotion choices (e.g. more restrictive city raider force promotion (if we are weaker, etc.), add optimal naval unitais promotion support, move some / a few settings to sas defines, add support for combat promotions and/or such and other fixes and tweaks anyways etc)
• fix/address AI settlers staying parked for a very long time in cities when it should be safe enough or urgent enough not to do so (e.g. 3 cities at turn 100 vs 1 city at turn 100 in some cases which was critically bad)
• AI now favours a military instructor specialist rather than a great general leader, as this is much stronger and scales better with late game: added logic to favour decremntially top 3 hammer cities especially more, and especially more so if they have or are building the heroic epic effect building. Also removed the limit on military instructors per city. Should lead to much stronger AI.

- Heavily reduced the cost of most wonders (e.g. 275->225, 525->425, 900->625-650, etc.): they were too costly to be significant strategic alternatives to just unit spamming. Hopefully should reduce early bankruptcies and lead to a more efficient economy and faster tech pace.
- Multiple other balance changes (non-exhaustive):

• increase TECH_DEPOPULATION 's negative health + happiness effect from -6 to -8. Adjusted and rebalanced the health and happiness gains from other future era techs. The total is +8 / +8 health / happiness gain. Moved effects so that they mostly apply in later techs to increase their strategical importance. Also Adjust tech flavors and such if any other such but anyways etc to match these changes anyways etc
• Barbarian civ-specific buildings can be captured (`iConquestProb`) with half the chance of their generic building equivalent (since their cities are overall easier and lower risk to capture, it balances it out). See Sevopedia/XML.
• BUILDING_PUBLIC_TRANSPORTATION 's iUnhealthyPopulationModifier effect from -15 to -20, increase cost from 150 to 200
• swap BUILDING_PENTAGON and BUILDING_EIFFEL_TOWER tech prereqs to be more historically accurate in the tech tree timeline,
• move BUILDING_THREE_GORGES_DAM to TECH_MARINE_TECHNOLOGY

- (Much nicer) Bonus specific builds logic is now fully customizable in sas defines, and refactor/enhance code anyways etc
- fix/tweak incorrect sevopedia ice cap information, and of terrain peak or/and such 's unit(s) panels information as well
- moved more tunables to sas defines
- multiple performance and code/refactoring optimizations
- 48 civs DLL included to match these changes
- update docs and other tweaks and fixes

Recommended to upgrade to this new version if you are using 5055 or an older version.

see for full changes: commit history from AdvCiv-SAS 5055 to 5082 viewable at https://github.com/wonderingabout/AdvCiv-SAS/commits/tech-rework/?since=2025-10-11&until=2025-10-31 or at https://github.com/wonderingabout/A...74...37615028018c14723ab4598cf095c39604f2a002


As for previous mentionned change ideas, i can't promise i would do them, ideally would but not sure, i may or may not.