• Civilization 7 has been announced. For more info please check the forum here .

[MOD] More Naval AI

- FreeTacos, why do you want Pirate coves on fishes: they are already too good!

Sometimes, when playing on medium and small with island chains, or archipelago, or even erebus continent with high seas and low cohesion, you just have to cover a bonus to make the whole pirate cove network fit. I figure, you can put an workshop or cottage or farm on any grassland as long as it doesnt have mana or a unique feature, so what the heck...

Besides, Pirates wouldn't avoid a spot just cause there happened to be extra food there!

I usually won't do it just to squeeze in 1 cove, as more often than not, I would have to add a marginal city to cover it anyway.

So I don't do it much, but now I don't have to use world builder.
:goodjob:
 
You bug replication directions are a little confusing. The scout starts 1 spot east of the fort, and cannot move to the NW of it in one turn due to terrain. I tried moving N of it, NE of it, SW of it. I played several turns with several different variations and never saw a python error. Do you remember what it said?

I remembered a workaround.
 

Attachments

  • Capture.JPG
    Capture.JPG
    135.3 KB · Views: 160
I think the problem lies with the automated workers, I believe that as the AI cities change their focus, they change the improvements.

I simply build whatever it takes to find that high growth/production city, usually a combination of farms and windmills, mixed in with various towns, workshops, and mines. My guess is the AI sometimes builds too many towns (and/or workshops) at the expense of growth (therefore production & commerce as a result of low population), which is ONE problem. The 2ND problem may be related to civics, such as converting to Aristocracy without Agrarianism (especially on Hell terrain without Sacrifice the Weak). As one can imagine without the AI knowing how to maximize city growth, let alone plot upgrading (which I doubt the AI does much of), this can lead to weaker civs later in the game. Once I reach a point where my Civ can hold his own against assaults with a strong technology base, it's usually game over and it's hard to finish a game.
 
I took a look at MoM and RifE with the ErebusContients mapscript and noticed kelp for the first time.
Is it intentional that it doesn't appear in More Naval AI or is there a problem on my end?

Thanks
 
It is an option available on the ErebusContinent mapscript along with City Ruins and Graveyards which I do see. Just want to make sure I'm not missing something. No worries really.
 
on the first turn, after exploring Bradeline's well, I got the golden age event. For some reason I got to keep the golden age even after regenerating the world map...Bug?

(excellent modmod Tholal BTW)
 

Attachments

  • les hippus2.CivBeyondSwordSave
    71.7 KB · Views: 55
I believe Kael stated that a lot of the code that is run in the map on creation, is not executed when the regenerate map option is used and was not recommended for use on the Erebus map script. That was a long time ago and may have changed by now ? I am not sure.
 
That jives with my experience on both Erebus and Erebus Continents. I have gotten a lot of weird effects, including starting settlers ending up in the middle of the ocean, on regeneration.
 
I also have issues when regenerating Erebus mapscript. It was always here.
 
Pushing out a beta release of 2.0 for the holidays! I'm heading out of town tomorrow morning and wanted to get this out, even though I know it needs some work. I'm hoping you fine folks can identify some bugs for me to fix by the time I get back to town on Tuesday.

Installation Note: You will have to unzip and copy the folder found in the attached BUG Mod ini files.zip into your BTS directory under My Documents (same place you find the CivilizationIV.ini file). I haven't yet figured out how to make BUG create default ini files.

Version 2.0 Beta
Quick summary of additions.

  • BUG mod added (use the button in the upper left or hit CTRL-SHIFT-O to access options)
  • AI Autoplay gets an interface (hit CTRL-SHIFT-X)
  • New option added called Puppet States and Revolutions - combination of the Puppet States mod and the Revolution mod (note: this has not been fully integrated with FFH so expect some weirdness)
  • Introduction of CivilizationYieldRiverChange xml tag (thanks Terkhen!) - gives Illians +1 commerce on Ice River tiles
  • Fix for python errors reported by Broken Hawk
  • AI tweaks from Ruthless AI - mostly regarding diplomacy
  • A few more AI and Unofficial Patch tweaks found when merging the Revolutions mod
  • Possibly other stuff, I forget, it's been a busy past couple of weeks... most of the recent changes can be found on my Sourceforge page - http://sourceforge.net/projects/tholalsffhaimod/
 

Attachments

  • BUG Mod ini files.zip
    59.1 KB · Views: 88
2 problems so far:

~~ 2.0 breaks the scenarios for me, even if I load them directly.
~~The BUG alerts don't all work, I don't get notified when a city is going to become unhealthy or unhappy even though I have them checked.
 
We had an OOS in 1.8.5 (we still have not finished our last game). One of us is playing as the Infernals (we had the previously discussed OOS error while switching civilizations) and he is conquering his previous civilization (the Sheaim). Whenever he conquers one of the Sheaim cities, we have an OOS error.

EDIT1: The OOS happens at random when he conquers a city, not only when he conquers one of his old cities.
 
Regarding the previous beta patch:

The important promotions most of the time are shock, cover and formation. That is what allows a unit to be a counter to another. The AI seems to tend to get combat three before these promotions – but, for example, a combat one and shock warrior is the favorite against a combat two warrior.

Catapults and mages should never ever be alone.

The AI casts maelstrom and stuff on its own troops hurting itself and the enemy.

The AI builds too many useless buildings. Just because you got construction doesn't mean you should spam walls. Cities should have their functions and in consequence just certain needs.

All-out attacks leaving minimal defense is not a suitable tactic for the AI since it is not able to read the map and potential threats as a human can. They always need to leave some defense or potentially get screwed in several different ways.

The AI beelines stuff like archery and bronze working early on a little more now but they should typically get education before that or be too dependent on how good is the start. This is more for low levels of difficulty since on the higher levels they can beeline that pretty well (Although I wonder if on immortal or deity they should just beeline code of laws before anything else and just race away in technology (Well, cottages for them grow so fast on deity...))

The AI is not building enough units in the middle game once they are ready for war ( I can't believe I'm saying that). They should go crazy as long as it doesn't make them crash or slow them down technologically too much in a hopeless attrition war.

I played a game with Jonas on Monarch. I had 23 cities by turn 150. All built. Guess how many units I had by turn 200.

It's about time to tweak the damage caused by spells like Tsunami. All spells should be tweaked and it's damage reduced. There should be a limit on how many units can use a certain spell or at least a cooldown for a certain number of turns after using certain spells. Otherwise mages rule the game.

The AI should use collateral damage better.

The AI should be better at achieving building victories.

Also better in city placement and they should learn not to build settlers during war and settling them towards the enemy if they already have enough cities.
 
I suspect that the code found in the subversion repository is missing a file or a function. In the linking phase it is not able to find the definition of the CyCityPythonInterface2 function. I was not able to find it either.

Code:
   Creating library Debug\CvGameCoreDLL.lib and object Debug\CvGameCoreDLL.exp
CvDLLPython.obj : error LNK2019: unresolved external symbol "void __cdecl CyCityPythonInterface2(class boost::python::class_<class CyCity,struct boost::
python::detail::not_specified,struct boost::python::detail::not_specified,struct boost::python::detail::not_specified> &)" (?CyCityPythonInterface2@@YAX
AAV?$class_@VCyCity@@Unot_specified@detail@python@boost@@U2345@U2345@@python@boost@@@Z) referenced in function "void __cdecl DLLPublishToPython(void)" (
?DLLPublishToPython@@YAXXZ)
Debug\CvGameCoreDLL.dll : fatal error LNK1120: 1 unresolved externals
NMAKE : fatal error U1077: 'C:\Users\terkhen\Desktop\CivilizationDev\zzDev\TOOLKIT\bin\link.exe' : return code '0x460'

I suspected that a CyCityInterface2.cpp file was missing from the repository, so I checked RevolutionDCM source code for it and I managed to compile successfully after including this file in my source code:

http://revolutiondcm.svn.sourceforg...e/CvGameCoreDLL/CyCityInterface2.cpp?view=log

Did you make changes to this file? In that case, my merge will be quite wrong :p

EDIT: The game ran fine but it seems to be missing a lot of TXT_KEY_... keys. Maybe some XML files are missing from the repository too?
 
Top Bottom