PROJECT: Playtest (the 4th Thread)

[BureauBert]

I am glad to confirm that the AI diplomacy issues seem to be fixed in 597 -- I didn't notice any irregularities in this regard.
Congratulations and many thanks to our proud Source Code Team .

So far I didn't notice any other bugs either and I have no crash to report -- but I will continue crashtesting.

There is just one thing that is different from previous playtests (not all of them, but some): Turn processing lags (on my ultragig map, with a lot of SLIC, as you already know). These lags might be caused by some SLIC-processing or simply by mapsize and number of "game materials", but as they didn't occur with some of the previous builds I think it's worth reporting them -- maybe someone has an idea about what could possibly be done (or has been done and could possibly cause that difference). These lags start occuring about turn 80. They affect particular AI players for a couple of turns (2, 3, ... 5 turns), which *really* take a long time (it feels like the old "goody hut bug" where figuring out which advances/units are available could take several minutes). After that couple of turns that particular AI "recovers", and another AI eventually gets affected by this mental blockade. Usually this affects only one AI at a time, while all other AI's are running through their turns pretty quickly, although basically the same scripted actions need to be calculated at least for more than one AI and although all AI's are more or less equal in empire size and number of units.

Well, I guess if no one does ever notice this issue with the unmodded game, it must be due to my modifications . But even in this case it would still be interesting to figure out what needs to be done or avoided in SLIC scripting or what could possibly be optimized in the SLIC engine or something in order to get rid of those lags / to enhance the performance of a heavily modified game .

[Ekmek]

Originally posted by BureauBert
These lags start occuring about turn 80. They affect particular AI players for a couple of turns (2, 3, ... 5 turns), which *really* take a long time (it feels like the old "goody hut bug" where figuring out which advances/units are available could take several minutes). After that couple of turns that particular AI "recovers", and another AI eventually gets affected by this mental blockade. Usually this affects only one AI at a time, while all other AI's are running through their turns pretty quickly, although basically the same scripted actions need to be calculated at least for more than one AI and although all AI's are more or less equal in empire size and number of units.

I only tested it as the barbarians to turn 200 on a small map with 6 players just to test the diplomacy and didn't notice a lag. However, your lag sounds more like AI strategy processing. We faced that problem before wit the governor AI and unit pathing. It might have been introduced because there have been some governor AI changes between 589 and 597. The other possibility is that your slic is adding to AI's decision making slowing it down.

But I am happy to see there are no bugs so far. thats good since it may be 597 for a long time.

[Protra3211]

Hi E!

Have the new downlode 597 with playtest ,tiles -new EDiplomad.
No more problems with diplomacy! Thank you !
Well only played 117 turns but it has already gotten intense. Carthaginian and Malian are both spoiling for a fight with me. But for now its only a cold war build up.
One nation(Spanish) seems at least for now easy to deal with.
Im playing on Hard level 8 Civs.
Some background notes. Changed DiffBD numbers to match Apolyton Pack mod Hard level. In Goverment text gave each gov 5 more cites. Helps the AI.
I did have a small problem with the new EDiplomod. It seems that the file in English-Gamedata-Civ text had somthing missing to Jamica city strings.Game CTD and gave me the information. So I took your build 597 looked up that same file and by hand copied only the Jamica city string names to the new EDiplomod.
Worked without any more problems.

[Ekmek]

Originally posted by Protra3211
I did have a small problem with the new EDiplomod. It seems that the file in English-Gamedata-Civ text had somthing missing to Jamica city strings.Game CTD and gave me the information. So I took your build 597 looked up that same file and by hand copied only the Jamica city string names to the new EDiplomod.
Worked without any more problems.

Ah, man. Is it the number of cities (missing #s 21 - 30)? I'll fix it. If its something else let me know! If you noticed the original game had Jamiaca tags but text for poland so I splitthem, making an actual Jamaica and a Poland. You'll also see alot of extra strings in there so it'll be compatible with stuff from AoM.

please post your DiffDB numbers. I've noticed many playtesters go straight to the hard level so I think that should be the new medium. i've actually made a new DiffDB you can test it hear. It adds my AI no deficit stuff to the harder levels. But I'd like to balance it with yours.

[Protra3211]

Thanks E for the file.

I think I will do three levels in DiffDB. Also the Const file has numbers that are settings for generating the map.The Apolyton Pack numbers found in Const text file gives the best map. Less clumby and AI has easy time expamnding.
Back to playtesting -will give your file a try in next game

[Fugi the Great]

Loaded 597, and started running it and it crashed to desktop. Tried again - same thing. All files in archive overwrote existing files, no mods used.
Crash.txt:

0x005177fa [?GetTransType@UnitRecord@@QBE_NAAJ@Z + 0xa]
0x00517463 [?Create@UnitPool@@QAE?AVUnit@@JJABVMapPoint@@V2@PA VUnitActor@@@Z + 0x60]
0x004ddabc [?CreateUnit@Player@@QAE?AVUnit@@JABVMapPoint@@V2@H W4CAUSE_NEW_ARMY@@@Z + 0x332]
0x0052944f [?gameinit_PlaceInitalUnits@@YAJJQAVMapPoint@@@Z + 0x23c]
0x0052f496 [?gameinit_Initialize@@YAJJJAAVCivArchive@@@Z + 0x2e7b]
0x0040d71a [?InitializeGame@CivApp@@QAEJAAVCivArchive@@@Z + 0x196]
0x0040f18c [?StartGame@CivApp@@QAEJXZ + 0x11]
0x0040ffcf [?Execute@StartGameAction@@UAEXPAVaui_Control@@KK@Z + 0x12]
0x006e73a1 [?HandleActions@aui_UI@@QAEXXZ + 0x5f]
0x006e732c [?Process@aui_UI@@UAE?AW4AUI_ERRCODE@@XZ + 0x47]
0x0040ecca [?ProcessUI@CivApp@@QAEJKAAK@Z + 0x1fd]
0x0040f01a [?Process@CivApp@@QAEJXZ + 0xa7]
0x00407e01 [?CivMain@@YGHPAUHINSTANCE__@@0PADH@Z + 0x3b6]
0x00407878 [WinMain@16 + 0x74]
0x008bb8dc [WinMainCRTStartup + 0x134]
0x77e814c7 [__onexitbegin + 0x772c608f]

[Ekmek]

Fugi, could you post your game where it crashed? Or did this crash at start up? If it did you might have to do a fresh install of ctp2, the patch, the playtest, the build. Sorry but I haven't seen this crash before and I think it might be a problem elsewhere. But if you do have a savegame I might be able to find it (although transportation hasn't been messed with AFAIK).

[Tamerlin]

I have installed Rev597 over the 2005.10.23 Playtest and I get the following error message when I try running the game :

[MESSAGE]
APOL_const.txt line 225: Error

Expected keword POWER_POINTS_TO_MATERIALS not found

Continue?
[/MESSAGE]


It seems to be useful, so here is the crash.txt content:

0x00533d64 [?Serialize@SlicEngine@@QAEXAAVCivArchive@@@Z + 0x2a4]
0x005335ce [??0SlicEngine@@QAE@AAVCivArchive@@@Z + 0x39e]
0x004f886e [?gameinit_Initialize@@YAJJJAAVCivArchive@@@Z + 0x8fe]
0x0040b398 [?InitializeGame@CivApp@@QAEJAAVCivArchive@@@Z + 0x178]
0x0051bac8 [?Restore@GameFile@@QAEKPBD@Z + 0x718]
0x0051a5ab [?RestoreGame@GameFile@@SAXPBD@Z + 0x1b]
0x0040cbbe [?LoadSavedGame@CivApp@@QAEJPAD@Z + 0x11e]
0x0040d565 [?Execute@LoadSaveGameAction@@UAAXPAVaui_Control@@K K@Z + 0x15]
0x005fc839 [?Process@aui_UI@@UAE?AW4AUI_ERRCODE@@XZ + 0x139]
0x0040c480 [?ProcessUI@CivApp@@QAEJKAAK@Z + 0x1a0]
0x0040c882 [?Process@CivApp@@QAEJXZ + 0x112]
0x00407608 [?CivWinMain@@YGHPAUHINSTANCE__@@0PADH@Z + 0x3a8]
0x004070c5 [WinMain@16 + 0x65]
0x00792eec [WinMainCRTStartup + 0x134]
0x77e614c7 [__onexitbegin + 0x7736d3ef]

[Maquiladora]

Using Rev597 over 23rd Oct 05 playtest.

I get a popup whenever i enter a ruin with a barbarian in it (attached). Appears to be some alert related to the barbs?

Everything else is fine so far at turn 200.

[Ekmek]

@ Maq
It might be that you still have debugslic in userprofile set to yes. for some reason the debugslic line throws up a lot of slic errors

I didn't update the script.slc in my build.rar and it has the unnecessary text

alertbox '84BesetByVandals' {
	Text(ID_BESET_BY_VANDALS) ;
	DeclareWar(g.player, player[0]);  //EMOD
	Button(ID_BUTTON_CLOSE) {
		Kill();
	}
	UseDirector();
}

I had the declarewar line in there for barbs and its no longer necessary so if you add the '//' before the declarewar line it should get rid of it.

@TAMERLIN
This might also be the debugslic but I looked through my apol_const in the zip I have and it has this line at 225

POWER_POINTS_TO_MATERIALS 1  # 1 power point can be converted to x materials  (unused)

Does your have that line? It might be the slic in the mod your using I seem to remember one mod used the powerpoints stuff for upgrading or something. Note most of apolyton pack is part of the AE try giving it a try with out the apol pack.

[Ekmek]

@ Fugi

I think I found out what the issue is. I set my number of players to 12 (in the game) 14 max players and my player color to 2 and got your crash before the game starts. Messing with it and I saw that 11 and below there is no crash but 12 and higher. Since I usually play around that I caught it (but my tests are usualy 6 and below).
So I'll see where the crash is.

[Ekmek]

Grrr I got this crash and seems related to net play although i was playing single player. And I did notice the player lag bureaubert talked about. I was playing my civ3mod but it seems similar to the Fugi crash with 577 he posted here.

0x0076b1e8 [?begin@?$list@PAVGame@NETFunc@@V?$allocator@PAVGam e@NETFunc@@@std@@@std@@QAE?AViterator@12@XZ + 0x18]
0x00884783 [?Rollback_All_Agents@Plan@@QAEJXZ + 0x42]
0x00887b96 [?Rollback_Matches_For_Goal@Scheduler@@IAEJABQAVGoa l@@@Z + 0xe4]
0x00886732 [?Match_Resources@Scheduler@@QAEX_N@Z + 0x2c3]
0x008ac246 [?GEVHookCallback@CtpAi_ProcessMatchesEvent@@EAE?AW 4GAME_EVENT_HOOK_DISPOSITION@@W4GAME_EVENT@@PAVGam eEventArgList@@@Z + 0xa3]
0x005c2fc6 [?Activate@GameEventHook@@QAE?AW4GAME_EVENT_ERR@@PA VGameEventArgList@@JAAJ@Z + 0x8e]
0x005c3c5a [?ActivateHook@GameEventManager@@QAE?AW4GAME_EVENT_ ERR@@W4GAME_EVENT@@PAVGameEventArgList@@JAAJ@Z + 0x3a]
0x005c18c4 [?Process@GameEvent@@QAE?AW4GAME_EVENT_ERR@@AAH@Z + 0x2f]
0x005c3ab8 [?ProcessHead@GameEventManager@@QAE?AW4GAME_EVENT_E RR@@XZ + 0x33]
0x005c3a5b [?Process@GameEventManager@@QAE?AW4GAME_EVENT_ERR@@ XZ + 0x7b]
0x0040f076 [?Process@CivApp@@QAEJXZ + 0x103]
0x00407e01 [?CivMain@@YGHPAUHINSTANCE__@@0PADH@Z + 0x3b6]
0x00407878 [WinMain@16 + 0x74]
0x008bb8dc [WinMainCRTStartup + 0x134]
0x7c816d4f [__onexitbegin + 0x7bc5b917]

a reload of the save game at turn 42 crashed again but not with the net line:

0x008847d6 [?Rollback_All_Agents@Plan@@QAEJXZ + 0x95]
0x0088794f [?Remove_Matches_For_Goal@Scheduler@@IAEXABQAVGoal@ @@Z + 0x7a]
0x008867fc [?Remove_Goal@Scheduler@@QAE?AViterator@?$list@U?$p air@JPAVGoal@@@std@@V?$allocator@U?$pair@JPAVGoal@ @@std@@@2@@std@@ABV234@@Z + 0x43]
0x00886eb9 [?Prioritize_Goals@Scheduler@@QAE_NXZ + 0x313]
0x008862cc [?Process_Goal_Changes@Scheduler@@QAE?AW4TIME_SLICE _STATE@1@XZ + 0x53]
0x008ac132 [?GEVHookCallback@CtpAi_BeginSchedulerEvent@@EAE?AW 4GAME_EVENT_HOOK_DISPOSITION@@W4GAME_EVENT@@PAVGam eEventArgList@@@Z + 0x8a]
0x005c2fc6 [?Activate@GameEventHook@@QAE?AW4GAME_EVENT_ERR@@PA VGameEventArgList@@JAAJ@Z + 0x8e]
0x005c3c5a [?ActivateHook@GameEventManager@@QAE?AW4GAME_EVENT_ ERR@@W4GAME_EVENT@@PAVGameEventArgList@@JAAJ@Z + 0x3a]
0x005c18c4 [?Process@GameEvent@@QAE?AW4GAME_EVENT_ERR@@AAH@Z + 0x2f]
0x005c3ab8 [?ProcessHead@GameEventManager@@QAE?AW4GAME_EVENT_E RR@@XZ + 0x33]
0x005c3a5b [?Process@GameEventManager@@QAE?AW4GAME_EVENT_ERR@@ XZ + 0x7b]
0x0040f076 [?Process@CivApp@@QAEJXZ + 0x103]
0x00407e01 [?CivMain@@YGHPAUHINSTANCE__@@0PADH@Z + 0x3b6]
0x00407878 [WinMain@16 + 0x74]
0x008bb8dc [WinMainCRTStartup + 0x134]
0x7c816d4f [__onexitbegin + 0x7bc5b917]

EDIT:
BREAKTHROUGH

Ok, I think what caused the crash was my barbariansspawnbarbarians code. I saw that when the barbs fortified arond a city they REALLY siege it. I'll post a screenshot, but the multiply like gremlins in a swimming pool. so I'm going to have to tamperthat. but it seems the AI got boxed in and couldn't do a goal or something after a certain point.

also removing it seemed to speed up the players too. Fugi and Bureaubert were you using that code? or had a lot of spawn events?

[BureauBert]

multiply like gremlins in a swimming pool

Sounds fun, but I don't use any units spawn code -- I am forcing AI units/armies to do things, though, like settling etc. ...

Anyways I am getting this crash with 597, too -- and I am already familiar with it:

0x0076b1e8 [?begin@?$list@PAVGame@NETFunc@@V?$allocator@PAVGam e@NETFunc@@@std@@@std@@QAE?AViterator@12@XZ + 0x18]
0x00884783 [?Rollback_All_Agents@Plan@@QAEJXZ + 0x42]
0x00887b96 [?Rollback_Matches_For_Goal@Scheduler@@IAEJABQAVGoa l@@@Z + 0xe4]
0x00886732 [?Match_Resources@Scheduler@@QAEX_N@Z + 0x2c3]
0x008ac246 [?GEVHookCallback@CtpAi_ProcessMatchesEvent@@EAE?AW 4GAME_EVENT_HOOK_DISPOSITION@@W4GAME_EVENT@@PAVGam eEventArgList@@@Z + 0xa3]
0x005c2fc6 [?Activate@GameEventHook@@QAE?AW4GAME_EVENT_ERR@@PA VGameEventArgList@@JAAJ@Z + 0x8e]
0x005c3c5a [?ActivateHook@GameEventManager@@QAE?AW4GAME_EVENT_ ERR@@W4GAME_EVENT@@PAVGameEventArgList@@JAAJ@Z + 0x3a]
0x005c18c4 [?Process@GameEvent@@QAE?AW4GAME_EVENT_ERR@@AAH@Z + 0x2f]
0x005c3ab8 [?ProcessHead@GameEventManager@@QAE?AW4GAME_EVENT_E RR@@XZ + 0x33]
0x005c3a5b [?Process@GameEventManager@@QAE?AW4GAME_EVENT_ERR@@ XZ + 0x7b]
0x0040f076 [?Process@CivApp@@QAEJXZ + 0x103]
0x00777ee0 [?CheckForEasterEggs@ChatWindow@@QAEHPAD@Z + 0x124]
0x00777d97 [?ChatCallback@ChatWindow@@SAXPAVaui_Control@@KKPAX @Z + 0x84]
0x006e2051 [?HitEnter@aui_TextField@@SAXPAUHWND__@@@Z + 0x52]
0x006e25fb [?TextFieldWindowProc@@YGJPAUHWND__@@IIJ@Z + 0x101]
0x77d18734 [__onexitbegin + 0x7715d2fc]
0x77d18816 [__onexitbegin + 0x7715d3de]
0x77d189cd [__onexitbegin + 0x7715d595]
0x77d196c7 [__onexitbegin + 0x7715e28f]
0x00407e57 [?CivMain@@YGHPAUHINSTANCE__@@0PADH@Z + 0x40c]
0x00407878 [WinMain@16 + 0x74]
0x008bb8dc [WinMainCRTStartup + 0x134]
0x7c816d4f [__onexitbegin + 0x7bc5b917]

I have no real idea what causes it, but maybe doing something/anything with AI units via SLIC causes the internal scheduler to loose track, somehow Just from the wording of the error messages I would imagine the program tries to recalculate the entire situation if something has changed -- or has been changed via SLIC??? Maybe this "rollback" procedure kicks in every time the program detects the execution of an army-related SLIC event handler -- I mean, being paranoid doesn't necessarily mean you are not persecuted .

Similarily I am totally clueless re. turn lags. Sometimes I had the impression it could be related to the amount of cross-water activities -- it seemed to occur earlier on maps with more islands where some AI did some impressive cross-water settling, BUT regularily these were NOT the ones with huge lags in their turn processing. Maybe it's indeed more likely related to the amount of SLIC and the amount of trouble it causes THE SCHEDULER .

[Protra3211]

Im on turn 235 with the playtest and build 597 and so far its been good .
In 5th place(tweaked up DiffDB numbers except tech)and its been tough as one Civ the Malians have sent in two stacks of units 12 and 8 next to my cities to threaten to take them. Soon as I gave them a gold bribe the stacks moved off. Assyriains were doing this two but are now busy at war with another nation. Every see this before E?.
Not to many AI,s going to war with each other. Which is good as a strong empire needs time to build up then expand its borders by war.
The AI is much improved!

[Ekmek]

Originally posted by Protra3211
Im on turn 235 with the playtest and build 597 and so far its been good .
In 5th place(tweaked up DiffDB numbers except tech)and its been tough as one Civ the Malians have sent in two stacks of units 12 and 8 next to my cities to threaten to take them. Soon as I gave them a gold bribe the stacks moved off. Assyriains were doing this two but are now busy at war with another nation. Every see this before E?.
Not to many AI,s going to war with each other. Which is good as a strong empire needs time to build up then expand its borders by war.
The AI is much improved!

that's awesome. I never got to the point where 12 stacks were a threat, I'd been beaten by 6-stackers and had to deal with a wave 3 and 4 stacks before.

I notice the AI war rate does depend on what civs are playing. I've seen early world wars and I've seen long peace periods too. your game sounds like its going great, you ought to throw some screenshots into a story thread... or maybe we should try a succession game.