Demo 8.1 Goals and Plan

[Mark_Everson]

Looks like a great job Laurent. Having maps of some size work, even if a bit slowly is clearly of significant importance. I'll check it out next weekend.

Originally posted by LDiCesare
If needed, I might do stuff to the economy too (MarketForCommodity is a bit expensive).

Many of the Econ methods could do with optimization, since they were just done by me to be as straighforward as such a complex system can be. Good luck with the profiling and tweaking of the worst offenders. When we get to putting more merchants in, they will be a problem with combinatorial explosions also IIRC.

[LDiCesare]

I finally managed to get a user interface up and running for the diplomacy. There are only a few clauses working for now, namely peace and money transfer.
I coded an evaluation of peace so the ai decides whether or not it's worthwhile. We'll see if it's any good, but in the Carthago Delenda est scenario, Hannibal would refuse peace whereas Scipio would accept (considering that Scipio sees Hannibal has a bigger army whereas Hannibal sees no opponent, but both see Hannibal can seize enemy land).
I will add evaluation of money transfer (should be straightforward), and of money over X turns.
I do not plan to give different personalities to the ai right now. It's pretty easy to see how an ai would be cautious or reckless, vindicative or not, but that will come much later, as refinements.
I now have to find out why an ai would want to go to war. This will require a new scenario, since currently all scenarios start with war already declared.

[Mark_Everson]

Hi Laurent:

Thanks for getting this going. If you need suggestions on how to handle evaluations for war/peace decisions, I could probably provide something of some use based on the old D4 code. I may have already provided you with that code/info, just can't remember.

[LDiCesare]

The formula I used for peace is an immediate gain formula. I consider the value of the terrain in which military influence is higher and consider that this terrain would be mine if the war continued. I remove the cost of the military I expect to lose if the war goes on (the expectation computation here is very rough). Additional info could be added, like trusting the opponent or not, and mostly manageing other wars and other fronts.

[Mark_Everson]

What you've got sounds good enough for now to me. Later we can get fancier.

Cya,

Mark

[alms66]

Originally posted by LDiCesare
I now have to find out why an ai would want to go to war. This will require a new scenario, since currently all scenarios start with war already declared.

Just use the Ancient Agean Scenario... that's one of the most "complete" scenarios in Clash - as in it has multiple civs of varying power - kind of like a real game. You'd just have to modify all the civs to be at peace from the start, which was something I was planning on doing anyway, once diplomacy was in.

[LDiCesare]

Yes alms, I can certainly change this one, it'll be better than adding a new one just for that.

[LDiCesare]

I updated the code a bit more.
Here are things I did:
AI:
-Checking ai siege and reinforcement code.
-Made the policies ai better when it starts far from optimum. (helps the social panel and the ai civs)

Military: -supply level affects movement rate

Basic diplomacy.

Social model:
Civil rights - lower military units morale
Private Property - has an effect on economic model
Slavery: Slave social classes/groups created (no ethnic discrimination yet)
-Changed Poor Welfare Feeling: Was too high. Multiplied by (1-PrivateProperty*(1-tax)) the probability.
-Changed separatism feeling to allow better effect of high province autonomy in fighting the nationalism part and increased civil rights effects when PAL is too high. The result allows manageable riots in socialTest with province autonomies around 0.8.
-Removed "Replace Ruler" Feeling. Too boring and hard to get feedback from.
-Rebels shouldn't be able to take control of territory.

Miscellaneous:
Allow movement by NUMPAD. Note that it was already possible with other keys - left, up, home and such, but not numpad.
-show civ capitals on map with different graphics from province capital.

Performance
Minimised ai map view memory usage.
Memory leak problem. Fixed. Lets all turns in alms'DoWC scenario be processed in around 3 seconds even on turn 130 on my comp instead of freezing around turn 40.

Incidentally, I ported my code for generating random maps to civIV and it works well there too. I also learnt how to generate rivers between squares in civIV and decided we shoulddefinitely model the between square rivers diferently in Clash in order to be able to code a river generation system without getting headaches (no NOfRiver/WOfRiver tags please).

[LDiCesare]

What is/was still planned for D8.1 and where I am:

Naval ai needs completing in order to get amphibious warfare.

Military (Laurent)
-refine city wall availability, cost and effect
--> I'd rather not refine that more because if things need be done, the combat model should be redone to a system like dominions2 where you can see the battle if you so desire. So I'll not touch this further for D8.1

technology (alms, Mark, Laurent)
more complete ancient Tech tree. As has been pointed out many technologies will need game functions before they are interesting to the player. Integrate technology changes better with economics and social models, and modify military model as needed. Take ten to twenty technologies that currently have no effects and give them game effects(0%)
- integrate activity tag with subtag weight/rank inside technology tag(20% - activity now has a weight instead of proportion so you don't have to do all the computations by hand, but it's not yet integrated)
I haven't touched that part yet. I'm unsure this needs doing right now.

Government and social models (Laurent, Mark, alms)
Give all government social policies meaning in terms of economic productivity, technological development rate, military effectiveness, and in other areas as appropriate:
--> Slaves to provide for production bonus.
-Slight bug with apply policy button + Precision of autonomy.
-Rebels should cause some really bad event when they take control of capital.

Graphics
-Overlay for borders instead of inner white lines.

Scenarios and bugs therein:
-Attila: Why do Romans in Rome desire independance or autonomy when they are Romans???
-Random populated: Needs checking for riots. A bit exaggerated. Although it's fun to see that you can conquer a province with 0 unit and it suddenly riots and creates many.
-Ancient Aegean: "ERROR: order does not exist "None", In GovtEconOrdersInfo, check xml" is outputed a lot of times.

The most important stuff though is certainly saving the game. So should I refine the points stated above in order to get D8.1 out, let it out as is, or start coding save? The main novelties in the current code wrt D8 are, from a gameplay standpoint, a more complete riots model, some diplomacy and naval ai.

[Mark_Everson]

Thanks for the update and all the great progress Laurent!

Personally, I think that it would be best to finish up your list above and release before taking on save games. From everything that I've heard, the save part will be a major effort in itself. The best way to try to recruit new blood, which the project desperately needs, is to release a demo. Of course as we saw with the last one, there's no guarantee that many will play it.

If you send me the code and tell me what you'd most like me to look at, I'll take it for a playtest.

What was your basic opinion of Civ 4? I haven't gotten it b/c I have heard of many problems with older systems like mine (2+ years old). I tried to d/l the demo to check system compatibility but the d/l bombed.

[Lord God Jinnai]

You need to advertise it well too.

[LDiCesare]

I've added a 0 to 10% production bonus for slavery (+10% at 100% setting). Slaves are a social class unto themselves, and don't like slavery, and are more numerous if you increase slavery, so it leads to them riotting if you put high levels of slavery.
I think I fixed all crashes in the naval ai code now.
On the things I still have to do on the D8.1 list, there mostly remains "more techs" but I think it's only interesting in long games, which requires saves first.
So I'm going to send the latest version to Mark (when I've finished all the reinstall and other nice stuff you sometimes have to deal with with a computer) for foolproofing and would like that to be a demo if he believes it's stable enough. I'll then freeze the features for a while and work on saves, then maybe graphics and the tech tree and effects.

[Mark_Everson]

Sounds good Laurent! Looking forward to getting the new version.

[LDiCesare]

I've uploaded the Demo 8.1 code (
http://clash.apolyton.net/demo8/Clash_D8.zip ) and updated a bit the dusty Demo8 download page ( http://clash.apolyton.net/demo8 ). If you have problems running the new demo, please send me a mail. For comments on the game, post in this thread or the Demo 8 thread ( http://apolyton.net/forums/showthrea...hreadid=127272 )

[Gary Thomas]

I have started looking at the code in order to implement save/restore. Sorry about the long gap.

Cheers