AND2 and SVN Bug Reports - A New Dawn 2 ONLY

Well, I assume you did more work than I just did on this, since you're mentioning what I assume are key "transition" techs, and I therefore assume that you're trying to match dates for these techs no matter the game speed. I'm still curious what you came up with, of course, and I'm hoping that you value my point on having 900- and 3600-turn speeds available, given the other speeds you mentioned existing.
 
I'd like to do one for 3600 turns, too, since to me that's a better point between 2400 and 4800 turns, although obviously you can keep 3000 turns as well. 900 and 1500 turns are other settings of interest given your range of numbers.

There is a 3600 turns game. I've mistakenly written 3000 turns instead of 3600 in my previous post. Gamespeed will be 600, 1200, 2400, 3600 and 4800 turns. I have an option for putting 3000 turns also, but maybe it's not so useful; same for 900 turns. So your suggestion of 600, 900, 1200, 1800, 2400, 3600 and 4800 is what I was also thinking about. :)

600 turn advancements: 20*100+70*50+75*20+65*10+60*5+70*2+60*1+60*6/12+60*3/12+60*1/12=8200; transitions at 4000 and 500 BC, and at 1000, 1650, 1950, 2090, 2150, 2180, and 2195 AD.
900 turn advancements: 90*50+95*20+90*10+100*5+110*2+95*1+100*6/12+100*3/12+120*1/12=8200; transitions at 1500 BC and at 400, 1300, 1800, 2020, 2115, 2165, and 2190 AD.
1200 turn advancements: 50*50+145*20+150*10+150*5+140*2+145*1+160*6/12+140*3/12+120*1/12=8200; transitions at 3500 and 600 BC, and at 900, 1650, 1930, 2075, 2155, and 2190 AD.
1500 turn advancements: 20*50+190*20+180*10+180*5+180*2+180*1+190*6/12+200*3/12+180*1/12=8200; transitions at 5000 and 1200 BC, and at 600, 1500, 1860, 2040, 2135, and 2185 AD.
1800 turn advancements: 200*20+220*10+230*5+215*2+225*1+230*6/12+240*3/12+240*1/12=8200; transitions at 2000 BC and at 200, 1350, 1780, 2005, 2120, and 2180 AD.
2400 turn advancements: 100*20+330*10+320*5+345*2+335*1+330*6/12+340*3/12+300*1/12=8200; transitions at 4000 and 700 BC, and at 900, 1590, 1925, 2090, and 2175 AD.
3000 turn advancements: 450*10+410*5+430*2+430*1+440*6/12+420*3/12+420*1/12=8200; transitions at 1500 BC and at 550, 1410, 1840, 2060, and 2165 AD.
3600 turn advancements: 350*10+520*5+550*2+540*1+560*6/12+540*3/12+540*1/12=8200; transitions at 2500 BC and at 100, 1200, 1740, 2020, and 2155 AD.
4800 turn advancements: 150*10+750*5+760*2+770*1+790*6/12+800*3/12+780*1/12=8200; transitions at 4500 and 750 BC, and at 770, 1540, 1935, and 2135 AD.

These turn length quantities were made to ensure constant "normal-looking" dates, only one outlier (the earliest one) in terms of quantities of turns at each turn length, and a low standard deviation for the non-outlier turn quantities (for regularity per turn length quantity after you pass the outlier).

Problem with turns lenght/number of turns is that to make the game balanced, you should also take into consideration that techs will cost more in advanced eras and that you produce more tech points in advanced era. If you're using tech diffusion, techs in advaced era spread faster also. So, unless you want to build your spaceship in, let's say, 1520AD, tweaking on eras xml and other files as well is necessary. Or, you could simply rework the cost of every tech, but that was a loooong way to face the problem. So I've choosen the other way. Anyway here's what I'm using right now and it seems to work pretty well (I think you won't find it extremely "good looking", but it's not so bad):

528*70+480*70+168*80+144*60+84*10+48*20+12*120+12*120+24*40+12*10 = 98400 months
which in years is
44*70+40*70+14*80+12*60+7*10+4*20+1*240+2*40+1*10= 8200 years **

so transitions are at 2920BC, 120BC, 1000AD, 1720AD, 1790AD, 1870AD, 2110AD, 2190AD, 2200AD. (of course these are NOT era transitions; they're just where months per turn change).

More or less, the same for other gamespeeds. The difference between your model and mine, is that, for example, given that Theology is usually discovered between turns 150 and 180 by AI, in your model Christianity is founded between 500BC and 1650AD; in my model, it's founded between 20AD and 440 AD; which is still pretty loose and approximate, but I think it's better. :cool:

And yes, I will try to introduce a new gamespeed (renaming them) at 900 turns. I was thinking about it and now that I see others are interested in it, I will do it :)

** Edit: 8200 years of course, not 8800 as I've written before
 
Well, I hope you pick our option of 600, 900, 1200, 1800, 2400, 3600, and 4800, then. ;p

The midway points of the game under my settings would be 1970, 1950, 1860, 1860, 1815, 1695, 1620, 1580, and 1510. (It goes backwards over time because I included one-month turns in the 600-turn game, while I have no replacement for the obsoleted turn lengths under slower speed settings.) If you're building a spaceship by the time the game is only 50% over in terms of turn count perhaps you could simply consider yourself skilled, lucky, or just "unrealistic" in the sense of your civilization having solid, single-minded leadership for 100% of its existence over thousands of years.

I'm not sure what you mean by "balanced". It seems to me that eventually you'll have any number of what I call "god cities" (cities that are just crazy good), and techs will just blaze along every few turns (or better) on eternity speed. Picking harsh map settings only delays the time at which this occurs due to needing more tech to make up for it. Of course when I say "you'll" I can only speak for myself, but balancing the game for someone else's tech rate will only imbalance it for me and many others. There are a lot of settings that really change things, such a tech trading settings, revolutions, map size and type/"quality", etc. I think there's little chance you'll actually get techs to come at the "right" time in my games, because I use some drastically relevant settings and not others, and I don't believe you're going to be able to constrain a runaway player (which is inevitable in my case even on deity difficulty, assuming I don't somehow get stuck with a mediocre share of the world while simultaneously not getting wiped out eventually; I suppose this is possible in an always peace game, but I doubt you're balancing for that) to your dates. Of course you can feel free to do whatever; my dates are just a curiosity based on those three factors that I mentioned determining them.

Incidentally, your 600-turn settings have time go by faster per turn than it was before for some reason. I really can't say I agree with that, no matter your reasoning. Civilization doesn't include dark ages for a reason, and given that RoM-AND goes past current technology it's hard to accept that for no reason technological development is forced to slow down (in date terms, of course). The future in RoM-AND does not appear dark from my perspective, anyway, so I don't understand why it "should" go by faster (per unit of "accomplishment"). I'd much rather have the date time spent per turn always be either equal to or less than the amount spent for the previous turn. If it ever slows down you have basic unexplainable "issues" such as your military suddenly taking longer to move around in spite of your tech and even (at least most of the time) economy ONLY increasing (alongside the fact that the units didn't suddenly downgrade, either).

Another thing to consider is that we've only run one simulation of our current technological timeline. If the ancient Greeks had invented the steam engine like they very well "could" have instead of staying lazy slavers we'd have skipped QUITE a few years of comparatively mediocre technological development. Similarly, if the Dark Ages didn't occur Europe would've gotten its head out of its ass and started teching more quickly a LOT sooner. I'm sure there are other paths history "could" have taken in various locations throughout the world that would've each shaved centuries or millennia off the historical rate. So the historical rate is not very comparable to a RoM-AND situation in any case, besides RoM-AND being impossible to "balance" for given all the game and map settings and game occurrences that can have massive effects on the tech rate.

Also, your system doesn't meet my standards for "normal-looking"ness. ;p

I always play on eternity, so when I next found Christianity (or fail to, but I try to pick up most religions; Kemetism and Naghualism aren't worth the effort, while I try and very typically succeed in taking Zoroastrianism as my actual religion) I'll tell you the turn count. I'm still using 1.75C, though.
 
Well, my idea is that it's fun to play every era of the game; so balancing the game means that human and AI players are fighting at the same level of techs. Which in AND2 wasn't working because AI was always falling behind in techs, unless you were playing on the hardest level. Then, if you are good enough at ending the game before modern era, either by conquest or other victory, that's fine with me; in my view, you loose the fun of playing through later eras, but that's fine. But if someone is bragging to have conquered the world by 1000AD after 150 turns, well, there's something wrong in how AI is playing or maybe that person should play at a harder level. The point is that it was way too easy to be far ahead of AI players in terms of techs (and military strenght).
As for "balancing", at 600 turns speed, you get a tech every 1-2 turns in modern-transhuman era. That's because if you don't get this pace, AI will never be able to discover every tech before 600th turn. If it doesn't, it will never have a chance to win because it means that it's teching too slow. But playing at eternitiy or any slow gamespeed, will make it slower to get techs. I don't understand why I should tech faster at these speeds. I suppose I will anyway need 8-15 turns per tech (if you're the tech leader) in transhuman era at eternity gamespeed, altought I haven't tested it yet. Which I think is good enough, but it can be made slower if necessary.
I know that balancing the game for someone else can be difficult, because gamestyles differs from player to player. That's another reason why it took me so long; but the main reason I'm doing this it's because AI was too weak/slow in teching: if in any game you can reach future tech while AI can't, then AI is teching too slow and hence you have a huge advantage which makes the game too easy: this is the main point.

For the very same reason, we can't explore with Civ what would happen if "ancient Greeks had invented the steam engine"; that's because there's a tech tree which IS based on our timeline technological development. And this has nothign to do with dates but with turns: I could simply rematch months-per-turn and have a civ discover Steam Engine in 5500BC as well as in 28456AD, but that would always be turn 330/600 (for example, in normal speed).


I'd much rather have the date time spent per turn always be either equal to or less than the amount spent for the previous turn.

I know, I was thinking about it too. I have done the opposite only occasionally for the transhuman era and only to make it so that the game has the right number of turns and ends in the correct year. I think there's no way to solve it if not by reworking again every xml regarding research, which I think it would take way too long.

As for the example of Christianity, thanks if you let me know about the turn in which it's founded (and year too, please); but ver 1.75C was very different from AND2 because techs/units/other things were removed or replaced.
 
Well, I just played a fair amount of an always peace 6-AI deity game, but my capital's location wasn't exactly amazing, and between that and population growth (I suppose I could've limited this, but I had no tiles in particular that a new city would give me that were worth trading the tiles I had plus the third city maintenance hike for) I never got particularly ahead enough to found a third city without ruining my tech rate below 60% (I toggle between 100% and 0% science in batches because otherwise you lose both science and gold; a fix for this would be nice, even if it was just a toggler you could set minimum gold levels for alongside proportions for other things (in culture's case you'd need a way to keep it fixed, though), along with a toggle for the toggler, of course), and of course the double-settler-starting AIs kept expanding with their cheat gold to support it, so it became a bit too gappy for me. I actually got to where I had nothing to put in the queue but warriors that I would disband for gold immediately, and this is on eternity with all the extra 1.75C buildings enabled. I wish there was a setting between immortal and deity, because I "quickly" pull ahead on the former, while the latter has that double settler start and so on, so there's a huge gap between them. I can try playing immortal teams, but that's a bit lame to me in some ways...

I don't like ending games via war either, as it's boring to me, and I enjoy managing non-war elements more. War just consumes production and gold and usually gives little in return, since I abandon most enemy cities, as they're typically both misplaced and too expensive for me to hold (since I'm keeping myself at my optimal number of cities regardless). Beyond that, I will beat the AI anyway, so it only serves to give me empty space (to fill in later and to make the AIs disfavor wars with me for now) or for revenge, which I don't really value...

I don't have a lot of experience with the transhuman era (starting in it gives me too many options to deal with at once, and reaching it requires my interest to remain in playing the same single RoM-AND game), but if it's at all comparable to the 3-6 turn tech rates I get on eternity in the "middle" of the game you'll have to triple the costs of all transhuman techs. Maybe you have it REALLY expensive already after 1.75C, but I'd "expect" to have even more cities (at least 50% more, if not 200% or 400% more) by then, and they'd have SO much stuff going for them, corporations included, that they'd just be ridiculously effective each, too, so I don't know. I see that Future Tech costs 31.25 times what Civil Service does, for example, so maybe that's enough to push me back to 8-15 turns per tech, but I doubt it would unless I was really unlucky with corporations or something. Of course balancing for me is probably stupid given that I don't find the game interesting at all below immortal and I use "weird" map and game settings, so whatever.

Incidentally, is there no way that building mines, cottages, and other things in forest without removing the forest will come back? I really like some features AND has, but RoM had the better tile improvement options, due to this and the fact that you didn't need to tech up like crazy to make cottages grow into something worthwhile. In 1.75C farming everything that can't be mined or isn't worth mining is my only real tile improvement strategy outside of taking resources. The cottage is just garbage for SO long, and it's the one you have to wait the most on (for it to grow) before it's worthwhile, too, which makes AND's hatred for it a double whammy.

My examples of the steam engine and dark ages are for comparison with the super-focused games that occur in RoM-AND. The fact that you're [going crazy with both skill and absolutely ever-present and persistent focus in a given mostly uncompromised direction] (everything in [] there does not apply well at all to reality) makes it completely reasonable from my perspective to beat reality by a millennium or two. Reality has lots of setbacks that just don't occur in RoM-AND. Wars don't actually delay things that much unless you're at war very often (or of course unless you don't do well), especially since a war gives you space to keep expanding in later, to keep improving your long term tech rate (70-80% is fun for a while, but you really want 60% and later, when you have about a dozen cities or more, 55% to optimize your expansion vs. current research (count culture and espionage as if they were gold, I guess), and expansion means a lot more research later, in the "medium" term).

Anyway, regarding Christianity, I play with tech trading and brokering off, so it wouldn't necessarily be as fast for me as it is for you. I don't really know at the moment. I think my map settings may offset my disallowance of all tech transfers, though, at least if my start wasn't bad. I guess things would be a lot easier if I turned off realistic culture spread, too...

Is there a good time to try out AND2? I don't really want to try it when the help files aren't accurate, at least, if that's an issue.
 
I don't have a lot of experience with the transhuman era (starting in it gives me too many options to deal with at once, and reaching it requires my interest to remain in playing the same single RoM-AND game), but if it's at all comparable to the 3-6 turn tech rates I get on eternity in the "middle" of the game you'll have to triple the costs of all transhuman techs. Maybe you have it REALLY expensive already after 1.75C, but I'd "expect" to have even more cities (at least 50% more, if not 200% or 400% more) by then, and they'd have SO much stuff going for them, corporations included, that they'd just be ridiculously effective each, too, so I don't know.

The further you go on with the timeline, the more the techs cost; that's always been like that; of course the further you go, the more science you produce, either because you have more buildings or more cities (or both), so it's just a matter of maintaining these two factors aligned with each other (science required and science produced). That's part of what I'm doing too. I've play-tested I guess something like 150 games in the last months (many with AI autoplay for testing purposes), and I think the result is good enough now.

Is there a good time to try out AND2? I don't really want to try it when the help files aren't accurate, at least, if that's an issue.

That's another thing to fix, because there are a lot of tips coming from RoM (before AND) which are now quite obsolete. I'm going through them all because I'd like to get rid of the useless ones and eventually introduce some new tip or modify some of them where necessary. But that's not on the priority list now. I guess the best time to try AND2 is anyway whenever you want; I usually learn from playing, not from reading the manual (or help), but that's a matter of preferences. :) (Although in the end I read the manual when I can't come up with the solution to a problem :p ).
 
Well, I'm sure it's easier to balance things for how the AIs typically act, but when I see an AI ally (yes, I sometimes play such games) prioritizing building wheelwrights VERY early in the game it really makes me doubt their relevance for humans...

I really prefer 100% accurate help files. It lets me plan ahead and confidently learn whatever I'd like to know. Without them I just get random surprises, and some of those can be after major investment (techs, buildings, etc.) towards a goal that the help files claimed I would get, making them quite undesirable to me. Incidentally, it would be cool if there was a full tech tree in there, too, just like there is a promotion tree. It's a bit annoying for me when I have to start a game to be able to see the tech tree. I know the starting techs and all, but it's really far harder to plan things out before or outside of games when I can only navigate the tech tree one tech at a time.

So do you have no thoughts on fixing the need to toggle between 100% and 0% science to avoid losing both science and gold ALL the time, or on making cottages worthwhile again outside of deep into the game, or on building more tile improvements without them removing forest, like it was in RoM? These things would be major improvements to me if "fixed" in some even halfway decent way. I know that for improvements it would require rebalancing, but I really do find it hard to justify anything but farm+mine+resource-gathering tile improvement spam. Cottages only exist for tiles that I can't farm or mine, and of course are created last. They're never something I actually want, because they're bad. In RoM or "plain" BTS cottages had worth to me, while it was still desirable to have at least some farms for various reasons.
 
So do you have no thoughts on fixing the need to toggle between 100% and 0% science to avoid losing both science and gold ALL the time, or on making cottages worthwhile again outside of deep into the game, or on building more tile improvements without them removing forest, like it was in RoM?

Not for the moment; the game looks to work good enough as it is. IIRC Cottages were "scaled down" because they were overpowered in some older version of RoM/AND. But possibly some improvement in this area will be done in the future. :)
By the way, I'm not sure I understand what problem you have with science. Are you saying that you switch from 100% science to 0% science because otherwise science is wasted? I usually play with multiple research, so that's not a problem for me, but I've never thought about this problem before, even when I wasn't using multiple research. Can you elaborate more on this, because maybe I haven't really understood what the problem is?
 
Speaking of the bugs listed in the first post; "Shields for cities or the Arcology building disabled" --> I couldn't find anything wrong with shields and arcologies: @IPEX-731BA5DD06, what was the problem exactly?
Also, I'm rebalancing tech diffusion as well; now most of the civs in modern era don't fall behind more than 6-10 techs (except those civs that only have 1 or 2 cities). I think that's acceptable.
 
Is there a way to patch in any OOS fixes that have been done to c2c. I think recently they just solved a bunch of OOS issues and it c2c includes a OOS log. Can this be used for ROM AND? I've been trying to play ROM AND with some friends but OOS issues are frequent.
 
45°38'N-13°47'E;11925175 said:
Not for the moment; the game looks to work good enough as it is. IIRC Cottages were "scaled down" because they were overpowered in some older version of RoM/AND. But possibly some improvement in this area will be done in the future. :)
By the way, I'm not sure I understand what problem you have with science. Are you saying that you switch from 100% science to 0% science because otherwise science is wasted? I usually play with multiple research, so that's not a problem for me, but I've never thought about this problem before, even when I wasn't using multiple research. Can you elaborate more on this, because maybe I haven't really understood what the problem is?
That's not what I mean at all (regarding multiple research). What I mean is how 17 commerce becomes 8 science and 8 gold if you go 50% on each, but is 17 science or gold at 100% for either, respectively. This loss is VERY common, and is done per city (is it not? even if it is not there is a loss, and it would make sense to simply hold partial numbers indefinitely for future turns (I think this would be the best fix, if it's possible)), so it's not a matter of just losing up to 1 science and up to 1 gold per turn, but up to 1 of each per city per turn (more if you throw espionage into the mix; only culture is worth producing every turn, assuming it's for the sake of happiness (otherwise it too is better off produced in bursts of 100% spending on it)). The average loss could be considered to be 0.5, so when you add the two up you end up losing 1 per city per turn if you don't toggle. There are actually more losses on average the further you get away from common multiples, as you are less likely to randomly match a loss-free multiple (unless of course you're going all or nothing like me) for any given one of your cities. If you have a multiple of 2 commerce you can run 50% science without losing anything (of course this is based off of the truncated end commerce, so pretty much any meaningful modifiers even on tiny amounts of commerce force you right back to all or nothing if you want to avoid losing science and gold ALL the time; more cities (with different commerce rates) or different modifiers for science and gold will quickly make it almost impossible to avoid going all or nothing to avoid losing science and gold). If you have alternative multiples of 4, 5, 10, or 20 (for all cities) you can also pick percentages that won't lose you anything. There are non-integer numbers in between that give other viable percentages, but in the end it's common even with one city to need to go all or nothing. The calculations you'd have to make to check it for multiple cities aren't worth the time when compared to calculating what to save up for 100% research in batches.

The AIs are also inhibited by these losses, so fixing this would make them play probably "around" 1% stronger in the long run, assuming your games get to where their cities produce over 100 commerce for large periods of the game. If not then it could be a boost of even 5%. Keep in mind that in true 4X fashion losses compound over time, so the fact that you missed that tech by one turn means that you started building an inferior building instead of the new one, which further slows you down in the future (now you miss a tech by two turns, then eventually three, five, ten, twenty, etc.), causing the opportunity cost of not going all or nothing to become ever more common. If you trade techs having a tech now instead of later can mean the difference between one new tech and two or even more new techs, too.

Okay, I tested a bit in my game where I currently have three cities. Sometimes it rounds 0.75 up and sometimes it doesn't, while 0.8 was always rounded up. So the problem is only 50-60% as bad as I thought (sometimes you might be able to actually gain with one city or with a bunch of luckily "synched" cities), which would give gains for "around" 0.5% up to possibly 3% when modifying the numbers I gave in previous paragraphs. The numbers didn't make sense when I rounded only after adding them up from all cities, so I think you do indeed have per city losses. I didn't test very deeply, by the way, so no one should expect me to know for sure under all commerce and slider rates.
 
Thanks. I'm sure my number of complaints is annoying, but it's only because I like RoM-AND, I swear! :o

Anyway, I suppose it isn't much of an issue with culture, as that's accumulated to the hundredth(?) of a point every turn. Although I imagine there would still be rounding issues, it's minor. Being a republic also discourages maxing taxes to the point that I come out behind overall if I do it in my current game. I don't really know well how espionage is dealt with.

City food and production and empire-wide research all overflow, so it would be nice if city science and city gold also did (I suppose an even better fix would be to only hold partial points on an empire-wide scale, rather than per city as I suggested before), so you could safely ignore the minmaxing I don't ignore early on in games, before civics or desired happiness from culture restrict you more.
 
So jtanner already left the project... not surprising.

AND will never be bug-free with good AI. Impossible to fix it. Just forget and leave to die.
 
So jtanner already left the project... not surprising.

AND will never be bug-free with good AI. Impossible to fix it. Just forget and leave to die.

Well, actually there still someone who's trying to fix it and improve it. As longs as there are people willing to contribute, it won't die. But if that's your opinion, you're more than welcome to leave this subforum and find some other mod which best suits you. :)
 
Have any out of sync bugs been fixed since Aforess' release?

Yes, some of them should have been fixed by jtanner28 before he left by merging C2C code in AND2, but I don't know exactly what he did before leaving. I'm planning to merge more code from C2C to fix OOS, but it will take some time. Anyway I've been able to play a full game with my wife without too many OOS (and none of them was fatal). :)
 
I have attached logs for a game which is experiencing out of sync errors.

Using Beta 2 build of AND.
 

Attachments

  • Logs_Client.zip
    1.4 MB · Views: 310
  • Logs_Host.zip
    1.3 MB · Views: 287
Top Bottom