Advanced Civ

[crullerdonut]

I just want to say that I appreciate both Long Try's playtesting and feedback, and f1rpo's thoughtful consideration and responses.

 

[f1rpo]

I've reproduced the 0% culture issue through WorldBuilder. Culture spread happens at the end of the human turn, the enclosed tile gets assigned to the human player at the end of the game turn and remains at 0 culture until the end of the next human turn. In BtS, it'll actually remain at 0% indefinitely. Anyway, I've changed the help text so that it shows 100% ownership when all civs have 0 culture points. Showing 0% is kind of more informative, but I guess that's not worth causing puzzlement; better to keep the enclosure mechanism hidden under the hood.

An ugly hotfix is better than a broken beauty, I suppose

For a quick (local) fix, you could try setting
self.W_SCREEN = 1600
(instead of 1024) in Assets\Python\Screens\CvForeignAdvisor.py. To move the screen farther to the left, the screen.centerX(0) needs to be replaced (with maybe sth. like 200, flat) in the setDimensions calls in CvForeignAdvisor.py and in CvExoticForeignAdvisor.py. Implementing even these simplistic changes somewhat properly, i.e. based on the available display width, could take a couple of hours. I was already going to give up on changing the horizontal position until I remembered, just now, that the "Exotic" script is involved too. The effort becomes more reasonable when implementing a bunch of related changes at once.

 

[Long try]

You're the man!
My latest game has assets locked, but I load an older save and it works like a charm! For those who use a game screen of 1366 x 768 like me (to get bigger fonts), the amounts are

self.W_SCREEN = 1366
screen.centerX(-171)

BTW, can I change the font size manually in XML or something? Then I'll switch to bigger screen and post the values. For now, things are great already: not only F4 "glance" looks better, other F4 tabs also improve with the exception of the 1st tab (which doesn't affect gameplay in any way). Not only that, it seems F1 and F6 benefit from the tinkering too, haha

 

[f1rpo]

BTW, can I change the font size manually in XML or something? Then I'll switch to bigger screen and post the values. For now, things are great already: not only F4 "glance" looks better, other F4 tabs also improve with the exception of the 1st tab (which doesn't affect gameplay in any way). Not only that, it seems F1 and F6 benefit from the tinkering too, haha

The domestic advisor and tech tree were already enlarged. (I also agree that all of them should be.) Font size can be changed in Resource\Civ4Theme_Common.thm (after the second "advc" comment in that file). I've made a lot of small tweaks here and there (shorter text or slightly wider widgets) to accommodate the current size, so, if they're increased further, depending on how much, some text will probably not quite fit or have awkward line breaks – though I'm sure that's better than having to squint all the time. And the city names on the main map can't be enlarged.

Huh?... Well, my case was to argue only for the worker's improvement speed, but now I learned that its and settler's production is x2.5, too. Any rationale behind them being units and x2.5?

Worker production cost is only x2. Making everything non-military extra expensive would seem more consistent to me – though I'm not sure that it was a bright idea of Firaxis to make units cheaper in the first place. I guess making Settlers cheaper relative to population growth would favor faster expansion in the early game.

Why trading a city can go so easily? If the German government decides to just give München to Austria, I don't think people there would just happily do the switch with smiles on their faces all day.

Only a portion of the culture gets converted, specifically 50% of the old owner's culture or 100% of the new owner's culture, whichever is smaller. To make the city more useful to the new owner. The government's decision having a pacifying effect seems plausible to me - though the rule arguably exaggerates this effect. BtS fwiw had been removing the old owner's culture entirely in the city tile and in the inner ring.

I wonder if it's only me or it applies to other players, but when dealing with the annoying cultural problems (partly listed up early in the post), the 1st tendency is to raze & resettle. If that fails, too, then they turn to a journey of total destruction.

I generally don't play Marathon if I can help it, but, when I did, recently, I found myself conquering a big swath of territory at one point, more than my units could've pacified, and, militarily, I could've kept going. In part, I attribute that to the difficulty level having been Monarch, but Marathon favoring warfare as much as it does is a factor too. Specifically warfare with a (big) tech lead and only a modestly sized army. Well, I don't see what razing would've gained me. Most of the cities can't flip back to their owner, they don't cost maintenance while under occupation. They do add to maintenance costs elsewhere ("from number of cities") and to civic upkeep, so I suppose, during the trudge toward an already assured Domination victory, razing could make sense economically, – but, then, Domination will require the land to be owned.

In other words, if you heavily nerf domination then people would just go conquest.

OK, so this is mainly a concern about that last stretch of the game? I've decreased the Domination threshold in large games, so that it's basically the same mechanism as for population: 25% greater than the next biggest civ. So it's hard to see how (scorched-earth) Conquest could become the easier route.

 

[Long try]

Font size can be changed in Resource\Civ4Theme_Common.thm (after the second "advc" comment in that file).

Wow, that's a lot of entries. Since I don't know which is which, I applied a global increase of +1 to everything. Doesn't break the game, but honestly I don't really recognize any change. Could you advise me the specific lines that affect the stuff that we see most on-screen? You know, scoreboard, unit info on the lower left, notice on top and under the tech bar every turn...

Another issue: I just began a game with Advanced Starting Points as Handicap. The value entered is 90154. And the mod calculated it as a starting date of 1700AD

While the average amount is 694, exactly as I wanted, it is just some dozen points above the Deity default of 651. But now I only have 720 turns to defeat 28 AIs instead of 1250! A bug, I suppose? It's weird to not have The Wheel nor Mining in 1700AD...

EDIT: Holy romanized fusil, I made my first moves and discovered that 1 of the guys has Musketman (and Longbow, no less) right on turn 0.

It's pretty clear that AI is equipped with all the 1700s techs while only I truly started in the ancient era. All 6 religions were founded right on T0, and Pacal offered me The Wheel and all kinds of techs in trading It's obviously a bug, right??

 

[f1rpo]

I can confirm that the unequal distribution of start points is broken. Good to know. I haven't figured out when that happened, but I think it's been that way for quite a while. A year ago, I noticed that unequal distributions cause a crash; I fixed that, but, evidently, that had not been the only issue. I know how to fix it in v1.09. Workaround for the time being: Not to configure unequal distributions. SPaH with a flat distribution should work.

Wow, that's a lot of entries. Since I don't know which is which, I applied a global increase of +1 to everything. Doesn't break the game, but honestly I don't really recognize any change. Could you advise me the specific lines that affect the stuff that we see most on-screen? You know, scoreboard, unit info on the lower left, notice on top and under the tech bar every turn...

These are the ones that I increased (quick screenshot from WinMerge):

Spoiler :

I don't know which ones apply where specifically.

 

[Long try]

Another quick one: the turn count for improvements on forest tiles aren't really correct. For example, if I pre-chop one for 2 turns, and then switch to "build a cottage", then it should display 1+4=5 turns. But in reality I still have to spend 7 turns.

 

[f1rpo]

I never noticed that, but you're right, and it's this way in BtS too. The Cottage is treated as a separate worker build entirely. I've started writing code for addressing this, but I think that gets too complicated, or at least too laborious. E.g. when switching from an incomplete Cottage on a Forest tile to a Farm, should any progress carry over? If more than 4 turns were spent on the Cottage, then apparently at least 1 turn was spent on removing the Forest. One could assume that the clearing always happens at the end of the build period, but wouldn't the worker have to make a bit of room for the Cottage first?

Getting back to the culture range subject, here's a screenshot showing the culture range of a Zulu city whose borders have expanded twice:

Spoiler :

Culture spread extends 5 tiles into Russian territory. The Coast tile straight south of the Zulu city is at 12% Zulu nationality. If I wanted to reduce that on account of Russian tile ownership along the way, I could do it based on the (air) distance to the nearest Zulu-owned tile. (Whereas enumerating all plausible paths of cultural transmissions seems like a way madness lies.) For gameplay purposes, I think it's helpful to spread culture this far, i.e. not just to the two border cities but also to Moscow at the edge of the Zulu culture range and into one half of Moscow's workable tiles – because that'll make it easier to assimilate Moscow if the Zulu manage to conquer it. I.e. not just the border cities are somewhat easy targets, things get progressively tougher farther into the rival's territory. In terms of realism, if one thinks of city tile culture as nationality (as the city screens labels it), then it does seem strange to encounter a minority of several percentage points in any city that isn't directly at the border. Maybe one can interpret it differently. I don't think anyone has raised this as a problem before. I have seen K-Mod criticized for the blur of colors on the Culture layer, but I understood that to be a visual problem, and I've (long ago) adjusted the thresholds for coloring a tile to address that.

 

[Long try]

when switching from an incomplete Cottage on a Forest tile to a Farm, should any progress carry over? If more than 4 turns were spent on the Cottage, then apparently at least 1 turn was spent on removing the Forest. One could assume that the clearing always happens at the end of the build period, but wouldn't the worker have to make a bit of room for the Cottage first?

Cottage to Farm is totally different builds, so no progress carry-over, I guess. Things would be much clearer (and I hope easier) if the game just treat all "build something on a jungle/forest" improvement as a combination of separate actions, chopping and then build. So an order on the surface is simply a sequence of 2 works internally, which means the forest always get cleared after 3 turns. That would appeal to your reasoning of "making room for Cottage 1st", which I totally agree for realistic reasons.

Culture spread extends 5 tiles into Russian territory. The Coast tile straight south of the Zulu city is at 12% Zulu nationality.

That's clearly a "uncontested border + 3" scheme. It's a bit incompatible with the fact that a tile's visibility is only +2 max. Moreover, the fact that the furthest aka weakest influenced tile still have 12% Zulu with only 2 expansions is absurd IMO. With 3 Russian cities exerting power there (1 of them capital!!), a reasonable amount should be hovering around 1%; 2% max. I think my call for cutting in half both the reach and the strength of culture is quite convincing with this example.

 

[f1rpo]

Cottage to Farm is totally different builds, so no progress carry-over, I guess.

We could also say incomplete Cottage into Chop. Would be rather unexpected to me to find that I can chop in 1 turn after working on the Cottage for 2 turns. Less unexpected after 6 turns spent on the Cottage.

Things would be much clearer (and I hope easier) if the game just treat all "build something on a jungle/forest" improvement as a combination of separate actions, chopping and then build.

Agreed. Of course they can't be entirely separate because the Forest must remain on the map. (Though I guess one could make a case for actually removing the Forest after 3 turns; then the combined build action would just be a UI convenience.) Well, I'll be leaving it as it is (all of it).

On tile culture:

[...] It's a bit incompatible with the fact that a tile's visibility is only +2 max.

A civ's religion can spread to fogged rival cities just fine. K-Mod had also added culture spread through trade routes; one of the first things I disabled because it was distracting without much gameplay value. But somewhat realistic and one wouldn't expect vision of the city to be a requirement for culture spread through trade.

Moreover, the fact that the furthest aka weakest influenced tile still have 12% Zulu with only 2 expansions is absurd IMO. With 3 Russian cities exerting power there (1 of them capital!!), a reasonable amount should be hovering around 1%; 2% max. [...]

It is pretty strange ... And perhaps 10% culture in some of Moscow's workable tiles is not really a big help when conquering the city. Might not matter much for game balance. So maybe reducing the amount of culture spread per turn based on the air distance to the nearest Zulu tile would be a (simple) improvement. That would also mean that culture spread will pick up as soon as Moscow is actually conquered.

 

[Long try]

Though I guess one could make a case for actually removing the Forest after 3 turns; then the combined build action would just be a UI convenience.

That's exactly the case I want to make, thanks for putting it much clearer There's no reason the forest must remain on the map until the end, as you pointed out earlier. Furthermore, that move would completely turn off any argument along the lines of "cottage to farm or whatever". Not to mention it'll solve the stupid turn count issue I raised in the 1st place.

So maybe reducing the amount of culture spread per turn based on the air distance to the nearest Zulu tile would be a (simple) improvement. That would also mean that culture spread will pick up as soon as Moscow is actually conquered.

Ahh, I think I understand the "air distance" stuff quite earlier now. Makes sense

 

[Jorunkun]

Hello all - just clocking in after a few months on the road to tell you I've been giving the latest version of advciv a spin on my usual settings (see sig).

After a few false starts, I've taken a game into the late industrial age now, running third behind a powerful-but-backward Ragnar and a small but very advanced Mansa. Great game, very challenging.

The pirate boat changes have worked out really well for the early game; for a huge map, the settings feel just right. You can fogbust them away in the early game but they remain a threat until all the coasts are covered and it still stings to get the occasionally two swordsmen dropped on yor doorstep.

The AI is performing really well, as are @f1rpo 's many balancing tweaks. As far as I'm concerned, this is the best civ there ever was. Kudos!

What I've noticed over the past year is that that certain world wonders - especially trade boosters like the Great Lighthouse - have a huge effect in the early game, to the point where getting GL early may be a dominant strategy and in need of nerfing. Generally, any wonder that scales multiple cities is relatively better on a huge map, whereas single-city wonders, including academies, are usually worth less, relatively, in an empire comprised of many cities. Can anybody think of an elegant way to address this, maybe make wonder effects scale with map size?

Because I play with random personalities on, I still get the occasional "favoriteciv" error after the age of astronomy, but the game runs stably. There are still memory leak issues with the flyout scoreboard, and it seems that some of the stats, like number of cities, don't show on it even though they are in my display string in BUG settings (!?WSZCJEPTUNBDRAHQ - it doesn't matter where I put the Q, the number doesn't show up). Anybody else get these?

Also, maybe because I use a 3840 x 2160 screen resolution, the info about active units in the lower left corner is sometimes positioned too low, the lowest line of text gets covered up by the interface UI frame. Happens with workers, specifically; see screenshot. Also, the intrerface and map font is very small on very high resolutions - any way to make that scale up in an in-game setting, maybe?

Lastly, what do you think of making the "delete unit" and maybe other rarely-needed buttons very small, so you (read: I) don't accidentally click on it dozens of times in every game? It probably looks goofy and is an inelegant solution, but I could live with it.

Just my two cents. Glad to be back.

 

[Leoreth]

Generally, any wonder that scales multiple cities is relatively better on a huge map, whereas single-city wonders, including academies, are usually worth less, relatively, in an empire comprised of many cities. Can anybody think of an elegant way to address this, maybe make wonder effects scale with map size?

Best solution I can think of that isn't "remove effects affecting all cities entirely" is to give all wonders affecting all cities a requirement of X buildings in your empire, like for national wonders. At least these requirements scale with map size and therefore make the wonder as a whole more expensive for large empires.

 

[f1rpo]

[...] There are still memory leak issues with the flyout scoreboard, [...]

I may have solved that problem at last. (Well, I don't think there's a memory leak involved. But the scoreboard strings should no longer get recalculated while the scoreboard is expanded.) I'll just attach my current state of the mod, let's call it version 1.08.1. Though I'm not sure if I should recommend it. Not really tested, not a lot of changes since v1.08 yet (though this also means less potential for newly introduced bugs). For a changelog, I can only refer to the Git history for now. Hopefully, I can address some of the other issues you mention in v1.09, sometime in the spring.

Edit4 (June): Never mind the links below; v1.09 has been uploaded with the latest bugfixes.
Edit: I've learnt that there's a new issue with the expanding scoreboard, showing info about unmet civs sometimes (only after having entered Debug mode?). Will fix a.s.a.p. - hopefully just a couple of lines Python in a Python script.
Edit2: Quick attempt at a fix attached (CvMainInterface.zip). To be unzipped into AdvCiv\Assets\Python\Screens.
Edit3: This still causes some delays. Better fix, probably (save link target):
AdvCiv\Assets\Python\Screens\CvMainInterface.py
AdvCiv\Assets\Python\BUG\Scoreboard.py

 

[Long try]

I've installed the 48DLL and cvmaininterface quickfix on 1.08, and it runs incredibly slowly. Does anyone experience the same thing?

 

[f1rpo]

AI turn times? Or the UI during the human turn? I've just run 150 turns on AI Auto Play on a Huge map with 20 civs. That took almost 6 minutes, which might be somewhat slow and, if so, could get much worse later in the game. 250 turns with 18 civs had only taken 5 minutes in my tests last year (with the 18-civ DLL). I'll have to run a proper test with the 18-civ DLL to be certain; so that I can compare with my old results. If the 18-civ DLL is as fast as it used to be, then there shouldn't be a significant difference for the 48-civ DLL either.

 

[Long try]

I think the new DLL in your latest post only works with 1.08.1. On 1.08, right after I loaded a saved game, everything slowed down like 20 times. The only moving thing I could see on screen is the circle running around an active unit, and it took 2 seconds for each small animation. So I installed 1.08.1 and tried to move all my stuff to the 'new' mod, and it works OK now

 

[f1rpo]

Yes, the DLL and Python changes are probably interdependent. I've noticed that my quick fix for the scoreboard still causes a delay upon moving the cursor onto or off the scoreboard. I think I can fix that too; for now, at least there seems to be no lag while leaving the cursor on the scoreboard.

 

[Spillsandstains]

might have been fixed, might be unimportant, but I ran the hoary old CivChecker on an installation of this and got:

UndefinedSymbol: XXL
   Used: AdvCiv gameinfo/civ4worldinfo.xml at line 225

Which is oddly what the tag on the back of my paisley y-fronts reads. Coincidence?

edit - you replied faster than I'd finished checking it out! OK all received. BTW thank you for your mod

 

[f1rpo]

Had to google y-fronts as well as paisley ... chic. Educational place, this forum.
I'd like to make CivChecker happy, but I don't want to add a TXT_KEY_... (saying "XXL") just to work around this false positive. The whole WorldInfo element for WORLDSIZE_XXL is normally commented out (I'd assume that CivChecker skips it then), I uncomment it only for stress-testing – though it can also make sense to use the XXL size when playing with the 48-civ DLL.