Bug Reports and Technical Issues

Thanks for pointing this out, but actually certain city sites use 20 as well, usually to mark settlement targets on faraway continents that are supposed to be prioritised. I am not sure if this extra scaling is necessary but I don't think it is an issue.
 
Playing as Chine on Monarch/Marathon game difficulty on almost the latest git commit.
Game gets stuck again when ending my turn. Deleting Kushan units on 100:52 solved the issue (just like before in my Persia game).
Savegame attached.I also have earlier savegames upon request.
Thanks again for submitting this save, it helped me identify the cause of this infinite loop and fix it. Unlike what I initially assumed, it was unrelated to the Kushans or their position on the continent boundary. Instead, the issue was trying to select a city defender from the units on (100, 52) where there is no dedicated defensive unit available. It repeatedly attempts to select the Skirmisher on the tile (whose deletion also circumvents the issue) but then fails to actually use it for that purpose, and then gets stuck in that situation forever.
 
Bad AI decision: AI Arabia get a Great Prophet in Al-Uqsur, and instead of using him to build the shrine for Islam (more widespread, his state religion), he uses it to build the Jewish shrine.
Glad that you submitted this because it actually revealed a more general bug that seemingly goes all the way back to the base game. Basically, the tries to figure out what the best city for a given wonder (or "limited building") and heavily discourages it in cities that aren't. However what it does not consider is when wonders can only be built in certain cities such as shrines. For now I disabled this logic for anything that is being constructed by a unit - because it already has logic in place to pick the best target city for its building anyway. After applying that fix, the Prophet actually goes to build a shrine in Mecca.

In practice this bug actually means that the AI would never have a preference for which shrine to build because they all would be considered to have negative value that is capped to 0. Perhaps in normal BtS your shrine city is usually a good enough city for this effect not to kick in, hard to say. That said I think it is a semi common observation in BtS game that a religion hoarding AI would sometimes make the strange choice to build the lesser spread religion's shrine first.
 
Playing as France on Monarch/Marathon game difficulty, when ending the turn I get an endless amount of python exceptions.
Savegame from the turn before attached.
 

Attachments

Playing as France on Monarch/Marathon game difficulty, when ending the turn I get an endless amount of python exceptions.
Savegame from the turn before attached.
Case of unfortunate timing - this issue was fixed in a later commit. If you want to update, your save is still compatible with it.
 
Japanese UHV allows for vassals - both explicitly stated and implemented.
You know, I had a feeling that might have been a possiblity, then remembered I failed a control southeast Asia goal because I vassalized Thailand. Turns out I was probably thinking of Burma.
 
Case of unfortunate timing - this issue was fixed in a later commit. If you want to update, your save is still compatible with it.
I tried with latest git commit today. It failed to load the savegame.
 
I linked the compatible commit that fixes the issue.
 
Still seeing Peru lose its UP on hitting the Renaissance in last Tuesday's commit. Not sure if some of the latest updates since its fix in July messed up its 'CvGameCoreDLL/CvEnums' value.
 
I believe you asked we keep an eye for that cheeky boi Mehmet pulling a fast one and "sneaking" into Italy. First file is autoplay turn before "illegal" army (in all cases army refers to the *small* Expansion groups - 2 Jans 1 Bomb) was noticed; I reran a few times and the same outcome was reproduced each time. Second file is the turn in which units have already spawned. Third save captures a *second* army spawn in Naples. Yet another army spawned in Naples a few turns later. You will no doubt take notice but I thought it noteworthy that during this entire time Ottos were only at war with Indies.

EDIT: For whatever this might be worth, I ran until Otto Expansion areas expired in 1450 and quite surprisingly they did not DoW any major civs. This was a pretty typical game state overall; normally they'd have attacked either Italy, Mongols, Egypt, or all of the above. I can't say this is definitely the first time ever witnessing such behavior but I think it is.
 

Attachments

Last edited:
Baibars leads the Ptolemaic Kingdom in 200 BC.
 

Attachments

  • Brennus BC-0200 Turn 150.CivBeyondSwordSave
    Brennus BC-0200 Turn 150.CivBeyondSwordSave
    478 KB · Views: 8
  • Civ4ScreenShot0002.JPG
    Civ4ScreenShot0002.JPG
    102.1 KB · Views: 20
I believe you asked we keep an eye for that cheeky boi Mehmet pulling a fast one and "sneaking" into Italy. First file is autoplay turn before "illegal" army (in all cases army refers to the *small* Expansion groups - 2 Jans 1 Bomb) was noticed; I reran a few times and the same outcome was reproduced each time. Second file is the turn in which units have already spawned. Third save captures a *second* army spawn in Naples. Yet another army spawned in Naples a few turns later. You will no doubt take notice but I thought it noteworthy that during this entire time Ottos were only at war with Indies.

EDIT: For whatever this might be worth, I ran until Otto Expansion areas expired in 1450 and quite surprisingly they did not DoW any major civs. This was a pretty typical game state overall; normally they'd have attacked either Italy, Mongols, Egypt, or all of the above. I can't say this is definitely the first time ever witnessing such behavior but I think it is.
Saves that are in the middle of autoplay are useless to me because it is impossible to interact with the game state.
 
Saves that are in the middle of autoplay are useless to me because it is impossible to interact with the game state.
Ahhhh right I took for granted nobody else knows my own personal method. Easy fix. Henceforth I'll just take more care to make sure I'm an active civ before saving. I usually do this by punching in the same year and getting switched to Harappa so as to affect the game as little as possible. I thought I did this hear, and am usually Paused before saving, so it might be the case of having to Unpause. That's just for those saves tho; none such complications in the future.

EDIT: Haaaa that explains that one Changelog item!
 
Ahhhh right I took for granted nobody else knows my own personal method. Easy fix. Henceforth I'll just take more care to make sure I'm an active civ before saving. I usually do this by punching in the same year and getting switched to Harappa so as to affect the game as little as possible. I thought I did this hear, and am usually Paused before saving, so it might be the case of having to Unpause. That's just for those saves tho; none such complications in the future.
In general observer mode is easier to handle for me than civilization autoplay, and can be used for the same purposes. You can start it with Ctrl+A and while there is no convenient way to abort it, you can do so with breakObserverMode() in the console, so that would be fine.

EDIT: Haaaa that explains that one Changelog item!
Yes, I managed to figure it out regardless, but it was a lot harder this way.

There is a mechanism that makes sure that when AI conqueror spawns target a city, they instead get created near the closest enemy city rather than the actual target. Otherwise it results in strange islands of conquerors that sometimes make no sense depending on how the enemy's territory is configured. Turns out that Naples is considered half way to Tripolis from Thessalonike. I considered extra conditions e.g. that the closest city should also be on the same continent, but eventually realised that doing this makes no sense for independents at all, who are just considered to be separate cities, so they should just be targeted directly.

Ressurected Ayyubids led by Rameses. Sorry I dont know when they did res this is just when I noticed.

EDIT: Later on they did change to Baibars.
Were they resurrected into the Classical era? Or medieval immediately?
 
Back
Top Bottom