<Nexus>
Traveler of the Multiverse
In general Civ4's AI understands xml changes. Unless you do something reeeeeally stupid, it will work. Simply put: whatever xml changes worked in your own mod should work in AdvCiv too 
Advanced Civ
- Thread starter f1rpo
- Start date
Right – by and large, the AI code should be as flexible as in BtS. It's possible that some unused abilities no longer work correctly, simply because they haven't been tested, but I might be able to fix that if I'm made aware.
--
Small update on naval blockades:
Thanks!Hello, amazing works, wonderful documentation !
--
Small update on naval blockades:
I don't think that's it. When I conquer a city that is blockaded by a war ally, the blockaded tiles aren't even visible to me. A savegame would obviously be helpful; or I'll probably come across the issue myself someday.
Elkad
Emperor
- Joined
- Mar 26, 2007
- Messages
- 1,057
I've got one, but two potential problems with it.
One, I don't know when it happened, so I just have a save with the broken blockade already in place.
Two, I have assets locked.
Should load right to the correct city. - Nagara Jayasri - assuming you can get it to load at all.
http://pages.suddenlink.net/flowerdog/images/Elkad-AD-1484.CivBeyondSwordSave
One, I don't know when it happened, so I just have a save with the broken blockade already in place.
Two, I have assets locked.
Should load right to the correct city. - Nagara Jayasri - assuming you can get it to load at all.
http://pages.suddenlink.net/flowerdog/images/Elkad-AD-1484.CivBeyondSwordSave
Last edited:
Hello,
I have something annoying with diplomacy : AI players don't stop spamming me with "oh no, we don't want this trade any more" but if I renegociate immediatly after that and ask them what do they want for the good, they aks me exactly the same trade they just broke !
Also, I have try nearly 12 or 15 regenerate map with moongose, and always end with a very very arid world with less then 10% grassland, little forest, and very huge plains and deserts. Is it normal ? It seem quite difficult to grow nice cities with that, and the good areas are so rares that even for starting position some player has a really poor area with no chance to compete. It seem I can't choice the aridity of the map, so I don't know : is it intented ?
I have something annoying with diplomacy : AI players don't stop spamming me with "oh no, we don't want this trade any more" but if I renegociate immediatly after that and ask them what do they want for the good, they aks me exactly the same trade they just broke !
Also, I have try nearly 12 or 15 regenerate map with moongose, and always end with a very very arid world with less then 10% grassland, little forest, and very huge plains and deserts. Is it normal ? It seem quite difficult to grow nice cities with that, and the good areas are so rares that even for starting position some player has a really poor area with no chance to compete. It seem I can't choice the aridity of the map, so I don't know : is it intented ?
Does yesterday's update help? There was one error in my code that had only occurred with the final-release DLL (fully optimized for speed) and not with the debug DLL that I use for testing. So the problem was probably worse than I realized, but might finally be -mostly- fixed now (probably not 100%).
As for PerfectMongoose – I'll check. Could be intentional or taken too far or broken.
@Elkad: Thanks. I've downloaded it. I'll try working around the locked assets.
I've refreshed my memory about the terrain and feature proportions generated by the PerfectMongoose (PM) script. I'm attaching two sample maps; just the first two I got with my current version of PM (I don't think it has changed since AdvCiv 0.96). I suppose yours look similar?
The overall desert ratio is as in Fractal, but the granularity is different. (And I've just noticed that there are very few desert rivers and thus few floodplains.) Fractal's ratio of grassland to plains is indeed much higher, about 11 to 5, whereas it's 1:1 in LunarMongoose's version of PM and 7:6 in mine. Also, plains and grassland form much larger clusters.
Forests are about as common as in Fractal - 16 to 20% of the land tiles in my tests. I think the distribution is also not that different from Fractal. They were quite a bit more common before I edited the script.
It would be nice if there was an option for the climate as most map scripts have that. The lack of options goes back to cephalo's original PerfectWorld script; see this post of his. It's easy to adjust the terrain and feature frequencies in PerfectMongoose.py under "GLOBAL CONTROL CONSTANTS". The granularity, though, is not adjustable. There are some things that I'd like to change about that, which seem impossible to do without learning how the terrain generator actually works.
Unfair starting locations are a major problem; I'm thinking that bulkier landmasses would help a lot (and would also be more Earth-like) – currently, there is often no fair way to distribute the civs over the many continents, and civs easily get cut off due to the jagged landshapes. So my plan is to experiment with the "legacy" land generator included in the script. It could be that the large tracts of deserts and plains also contribute to the problem. I find them important though as representations of the Earth's tropical deserts and steppe belts. Arguably, the steppes generated by PM are larger than they should be. I'd like to place more plains near the equator (and thus fewer plains elsewhere) representing clay-rich soils. I'm not sure if grassland should generally be much more common than plains, but, anyway, decreasing the "PlainsPercent" value in PerfectMongoose.py is easy to do.
Or you could download the Totestra map script, which – to my knowledge – doesn't use the latest and greatest algorithms from PerfectWorld 3, but has a great number of options. I don't expect that there'd be compatibility issues with AdvCiv.
Attachments
Hello, sorry to have not replied in awhile! You can ignore my prior report on relations. I've played quite a few more games and it was a non-issue. It just happened to be a game in which I never built up a high relationship. Subsequent games have been consistent with prior experiences, except that some of the modifiers take longer to reach max value.
Yes, I've been playing with my leaderhead changes for a long time. There are less wars, but watch out for the AI permanent alliances!
On another note, have you considered nerfing the early AI archer rush? It doesn't happen a lot, but it is very annoying. The AI can immediately start building huge stacks of archers. The only defense is to rush to archery or pray you have copper nearby. Of course, you never know if this is going to be needed or not because sometimes they build the stacks but don't attack.
Yes, I've been playing with my leaderhead changes for a long time. There are less wars, but watch out for the AI permanent alliances!
On another note, have you considered nerfing the early AI archer rush? It doesn't happen a lot, but it is very annoying. The AI can immediately start building huge stacks of archers. The only defense is to rush to archery or pray you have copper nearby. Of course, you never know if this is going to be needed or not because sometimes they build the stacks but don't attack.
Yes, it's ok now.
Yes, it is similar, but often deserts are bigger.
The fact is I found this world seem have already suffer from global warming lol
In -4000 BC, prehistorical world, grassland should be nearly all forested because the only areas were there is not forests are lacking something : rainfall, solar radiation, and/or temperature. Maybe the percentage of grassland is correct, but the result seems not because of the scale of cities as one Civ city cover the area of a small country. And also a very big "detail" : in real world we can export food wich make possible town in middle of nowhere like dubaï..
I'm attaching a screenshot from one of the savegames you sent me. I'd argue that Tokugawa didn't have any good options here – apart from the Horse, all the good resources are blocked by Jungle. Researching Horseback Riding, which takes forever, is pretty disastrous though. He really needs Iron Working and doesn't even have Mysticism yet for border expansion. (In case anyone wonders about the Salt resources and graphics – it's keldath's Dawn of the Overlords mod - based on AdvCiv.)
If that's based on the game with just Tokugawa on your continent, then I'm not surprised.
Attachments
Hmmm,
Thanks for the check.
So, how do you think i should tweak xml for improving the ai such as japan here?
Do you think the tech costs are too high on mine?
Oh, why no surprise? Suggestions on how to increase?
Thanks for the check.
So, how do you think i should tweak xml for improving the ai such as japan here?
Do you think the tech costs are too high on mine?
Oh, why no surprise? Suggestions on how to increase?
Last edited:
Well, Tokugawa is a notorious isolationist who won't trade anything unless Pleased and therefore doesn't ever become Pleased. (Unless you've already modified his XML values.) I have incomplete notes for a (discreet) overhaul of the AI personalities. My notes for Tokugawa are these:
Spoiler :
yeah the leaders are a good tweaking start.
once i had really aggressive settings for leaders.
archipelago - that the only map ill play
since civ2.
maybe ill find some way to increase map related bonuses.
and how about the low bards?
once i had really aggressive settings for leaders.
archipelago - that the only map ill play
since civ2.
maybe ill find some way to increase map related bonuses.
and how about the low bards?
Basically, the terrain generator picks the x driest plots and sets the terrain to desert. So there shouldn't be much variation in the total number of desert tiles. I did add a clause to prevent deserts in very cold climates (better represented as tundra or snow imo); that could lead to some variation, but the desert ratio among land tiles should still never exceed 20%. At least it's not supposed to happen.
Sure. This may not be a highly accurate map, but I think it gets the point across: slideplayer.com/slide/9449674/....jpg
Insofar, the higher forest density of the original PM script was more realistic. However, the whole representation of forests in the game is far from realistic and I wouldn't want early-game chopping to be an even bigger factor than with the stock map scripts. One can always argue that the forest feature represent only forests that are especially well-suited for construction (i.e. mostly hardwoods) and easily accessible.
Come to think of it though, I've aimed at the same ratio of forest tiles among land tiles as in Fractal, but Fractal has very little tundra (which is to say taiga), so that may not be the best point of reference.
I don't suppose there's one correct way to decide which soil types should be represented as grassland or as plains. I'll attach two more screenshots, both after 200 turns on Normal speed, one game on a PM map, the other on Fractal. I don't see a pronounced difference in the distribution of city sizes – I guess because the AI mostly avoids settling in the middle of the "brown" areas. In terms of realism, that's fine with me; Kazakhstan and Mongolia are sparsely populated to this day. As for gameplay, the PM screenshot suggests to me that you're right about terrain being a major source of unfairness. If Charlemagne had decent terrain to his east, he'd be able to compete. The coastline still looks like the bigger factor as it gives Suryavarman exclusive access to the northwestern portion of the map (at the expense of Tokugawa and Charlemagne). The terrain frequencies can be tweaked (and I intend to do so) and hopefully the shapes of the continents can be made more convex, but large clusters of similar terrain will remain (as they should imo), so it's hard to say if this can become a well-balanced map. Arguably, the algorithm that selects the starting locations would have to be smarter. As it is, it may only be suitable for role-playing and the Rise & Fall game option.
Edit: quote fixed
Attachments
Last edited:
Not sure about Barbarian cities. The low appearance rate of Barbarian units is probably also due to the map (fairly small starting continents – not much unowned land left by the time the Barbarians get going), but I've also noticed some artifacts from rounding in my code that could have a significant impact when continents are small. Fixing that could result in up to twice as many Barbarians.
By the way, I do intend to respond to all (or at least most) of the observations you've sent me in our private conversation.
Perhaps it's even still the same code that generates the landmasses (for almost all maps). I've been wondering about that; the code in CvFractal.cpp looks ancient – much more efficient than it would have to be for this millennium's hardware.archipelago - that the only map ill play
since civ2.
No dice. Apparently the same code in the EXE that prevents loading saves from other mods also enforces the "lock assets" option. And there seems to be some (subtle?) difference between your assets and AdvCiv 0.96 (and 0.96b/c). Well, like you said, with the lingering blockade already in place for some time, there probably wouldn't be much to see (and no debugger). Though if you do have screenshot ...
That's just fine. Thanks for the update.
In my (limited) experience, an Archer rush only happens at very close quarters with a warlike AI leader. So it shouldn't be that unpredictable and, with the default player counts, it probably shouldn't happen at all. "Should" ... Any details about the circumstances in your games?
Elkad
Emperor
- Joined
- Mar 26, 2007
- Messages
- 1,057
I've fiddled with the base game, and then moved the install folder to a new PC so many times I don't even know where original is any more. And I haven't seen my disks in at least a decade.
So short of sending you my entire civ folder (or at least every xml and ini file - since that's all I've messed with), I can't tell you what to fix to get past the locked assets.
Oh well, it'll happen again. Probably cognitive bias, but I seem to get it every 3-4 games (on huge/marathon). And I'm playing with assets unlocked.
So short of sending you my entire civ folder (or at least every xml and ini file - since that's all I've messed with), I can't tell you what to fix to get past the locked assets.
Oh well, it'll happen again. Probably cognitive bias, but I seem to get it every 3-4 games (on huge/marathon). And I'm playing with assets unlocked.
I've removed all civ starting techs (to help minimize ridiculous synergy starts, but free archery for AI), but give all players a starting worker. I'm assuming the AI is building even more archers than usual since they don't need a worker and can't build scouts yet. I've seen them build huge archer stacks early.
I also play with random personalities, so I guess I'm somewhat getting what I asked for with the unpredictability.
no worries mate.
another weekend , more of doto advc multiplayer time
i buffed up some xmls,
bonuses
barb handicap and some more.
more from my game:
ai is dead in the water. due to islands - no wars. and on the islands that have 2-3 - no wars.
strange thing though,
ai builds good defenses, archers, when theres longbow , cross bow and axeman.
another - extremely hard to take a city,
me and my friend , each in war with another ai,
when takingon a city - very hard to win, i had 1 enemy swordsman (why did he defend the city instead of an Axeman?)
and i lost 5++ troops over one unit, and that unit was barely scratched.
very hard to win combats, sometimes the strength value of the enemy will barely change after my unit had died.
Last edited:
