PROJECT: Playtest (Thread No. 5)

[Ekmek]

ok, i compiled it and I post a test build for the playtesters (in a sec)

[Ekmek]

LATEST PLAYTEST: September 9th 2006

2006.09.09 CTP2Playtest (11.8 MB)

LATEST BUILD REV 771 BETA: July 24th 2007


Revision 771 Beta (5.19 MB) (transport lag work around)
Edit: Removed links. For an up to date version of the Apolyton Edition click here.

################################

Beta testers this is the latest we have after the linux merge so that needs to be verified. also included are Martin's changes to difficulty and risk settings, its now a list box (so modders like Protra can add more). also included is CajunArson's work around for the transport lag we've all complained about. Let us (and him ) know how it goes.

sorry I didnt include a latest readme

[CajunArson]

Hey, I've tried out revision 771 and unfortunately there was no improvement in turn time.
However, I have a quick question about the build, I checked out revision 771 from SVN and the scheduler.cpp in that revision did not have my changes in it (it was the same as the 770 version). Are you sure the new code got compiled in?

Also, I'm curious about the debug statements in the code. Is there a way to get the output of these things by passing a flag to ctp2.exe, or do I actually have to get a debug build of the executable?
I'd like to investigate the problem further. Thanks again for your help.

[Ekmek]

Originally posted by CajunArson
Hey, I've tried out revision 771 and unfortunately there was no improvement in turn time.
However, I have a quick question about the build, I checked out revision 771 from SVN and the scheduler.cpp in that revision did not have my changes in it (it was the same as the 770 version). Are you sure the new code got compiled in?

I didn't commit your changes, i just renamed my scheduler.cpp and then put yours there.

Also, I'm curious about the debug statements in the code. Is there a way to get the output of these things by passing a flag to ctp2.exe, or do I actually have to get a debug build of the executable?
I'd like to investigate the problem further. Thanks again for your help.

I think you have to do a debug build. MSVC++ 6.0 allows for a debug compile of an exe, its just an option. Not sure what compiler your using. I'm not sure how passing a flag would work, Martin may be able to help.

Nice to see another coder! Many have come, sadly too few have stayed...

[Martin Gühmann]

Originally posted by CajunArson
Also, I'm curious about the debug statements in the code. Is there a way to get the output of these things by passing a flag to ctp2.exe, or do I actually have to get a debug build of the executable?
I'd like to investigate the problem further. Thanks again for your help.

If you mean those Assert and DPRINTF statements then you have to build a debug version. And of course all this extra crap blows up the size of the executible and slows down program and that in addition of the missing optimiyation.

Originally posted by CajunArson
I think you have to do a debug build. MSVC++ 6.0 allows for a debug compile of an exe, its just an option. Not sure what compiler your using.

Obviously he needs a windows based compiler, so how did you get your demo version of Visual Studio?

-Martin

[Protra3211]

Downloaded the new build and so far after 100 turns there has been no crashes.
Playing very hard level with production numbers from the impossible level. 10 civs with a 90 by 180 map.
Im already in 1st place.Been aggressive early as I built as many units as I could and have captured 4 cities.
This tactic works every time.In the mods the AI has a higher garrison count to keep this from happening.
Also have one wonder.Many civs are in a race for at least 4 wonders.
It seems the practice with the playtest has sharpend my skills.Like chess if your too passive all offensive options will be gone.The key is keep you units extended out as far as you dare so when declaring war yor within 2 to 3 turns from hitting your traget city.
In E.s Civ3 mod it was much harder to pull off early gambits like that as weak citeis with no units could be captured.But again good work the many added features make it enjoyable game,

[HuangShang]

Hey everyone I'm playing the newest build in the World of 28 scenario, which as always only has 13 civs. The bigger menus are going to take a little getting used to because my eyes don't jump to the right place yet, but I am happy at the changes.

I set it to very hard, and during my new strategy of making a mad dash to republic, I've been beaten to the Chichen, but only after I built Ramayana. The first wonder of the world was the pyramids. Right now I'm getting messages of Aristotle getting near completion, but I don't even have the advance for it. I also have the Great Wall and Forbidden City, as it seems that the AI does not like those too much. AFAIK all wonders get built pretty early so thats definitely fun.

Something more important is I still do not get more than 4 options for research, despite the new length of the menu. Perhaps it isn't enough just to make it long, because it's always had a scroll bar and they would rarely offer five choices.

Finally, my attempt at World of 27 crashes. I think it is on the teal USA turn, as you can see it crashes immediately in the USA save.

http://apolyton.net/upload/view.php?file=21215_Worldof27crash.rar

[Ekmek]

Huang you can change the number f advances in the const.txt, but i think your right the tech tree doesnt usually go above 4 (i just saw five in a quick test)

I check out your crash in the bit. Right now I'm really close to change the tribe select screen to a list box so we can have an unlimited number of civilization in a game. It runsfine but right now it:

- shows the barbarians as an option,
- shows all civs for a scenario
- doesn't update the leadername box on the tribe screen but it will show the name on the start screen.

So i'm real close (close than ever before!)


Thanks for testing

[HuangShang]

Wow I need to start opening these files. I never a trade route passing through a third city gives a little gold and that nukes only waste 5 tiles. Hehe, what are RAIL_LAUNCH_POLLUTION, PIRACY_KILLS_TRADER_CHANCE, and MONOPOLY_THRESHOLD (and how to have a monopoly)?

I just notice that Forbidden City is actually amazingly powerful, because the increased regard lets me threaten the AI into peace and tariff treaties. I have almost no units, just sitting quietly between Japan with one city in the Russian Far East and a very far away India. I remember impossible was much harder so I played very hard, but that Forbidden City is a wonder. Finally I can't figure out what research pact does.

[LemurMadness]

Woah... Transport lag workaround!

I am so testing this!

I currently don't have internet back home, Vista gave me troubles -- Constant freezes and lock-ups, maybe because the hardware was unsupported (although it was pretty new, an AM2 2.8Ghz FX2 system) -- So im back to XP x64 (At least we know it does works on Vista ).

We're getting near to a 777 build! It's considered a lucky number in the place I live. I'll give you guys feedback, I'll push the game to its limits all this weekend.

[Protra3211]

Thats great.A lot of action in my game as the AI is still pushing to reclaim 1st place.200 turns and not a single problem so far.I had one AI faction bring many stacks of units near my border.So I made a peace treaty with them because after a very long war with Japan most of my units are out of place.Some times diplomacy is the best strategy.Are you working on the AI in this new build 777?The AI is weak in defense of cities and is slow to build up forces and push to there borders like in the early builds.The hemming in tactic by the AI has always worked in the past as it keeps you busy with defense.
The AI does not have to beat you just keep you in check and their growth will over take you because of a higher score. Medmod 2 was noted for this.

[HuangShang]

Have to agree with Protra on AI actions. I started my game over on impossible this time, because I wanted to redo my cities placements really. So Japan is an island in Earth map, but they garrison 5 units per city at home, while they start settling the mainland and leave new cities undefended, but the first and much older mainland city they built is well defended with a catapult. It seems that they will maintain high numbers at home to prevent a backstab, but its worse if they built new cities at the front unprotected. I haven't gotten any longish turns, and the Japanese had definitely been shipping units. It isn't an assault though.

Back to wonders, I managed Ramayana right off the bat, and took Chichen as well soon after. I couldn't get any science wonders though, but I look forward to building myself an forbidden imperial palace again.

I rushed monarchy this time, heading to republic, and it works much better. The only noticable difficulty difference is a much increased contact with the rather far away Indians. I need to learn how to change the scenario in the future.

Finally the longbowman is a great addition to the game, fill a nice little niche.

[Protra3211]

I have found that using a map with 80 percent land with water ways around main areas works best.
There are many ponits where mods gave the game a tactical AI_Goals -Stategy files-force match -garrison count and diplomacy text file.Have made many notes in the past and have found some things that work best.
Have to do some research and try it out on the build.
Some kind of add on file to the build might be made.
The new builds have the AI managing there cities better than any of the mods as well as AI to AI diplomacy.Feels more like a multi game than a scripted AI.

[Ekmek]

any thing you guys learn/develop please post them so we can improve the AE

[demon59]

Trying to install 771b playtest. Fresh CTP2 install, then 2006.09.09 playtest, then 771b. Try to start CTP2 and get error message...

Error: Const.txt line 382
Expected keyword COMBAT_ELITE_CHANCE not found
Continue?

Clicking either Yes or No button dumps me to desktop.

Suggestions?
Am I missing something painfully obvious?