Cradle 1.3

[Spanscape]

Actually, the game I have crashing is only 8 civs and is on a medium size map. In addition, I have yet to conquer a single civ, and maxplayer is set to 0 while numPlayers is set to 9 (8 + 1 barb).

Playing these settings multiplayer with the Apolyton mod has yet to produce a crash in several games.

[stankarp]

I am certain it only happens on ultra-gigantic map.

Interesting that it only happens when the player saves, not during auto save. There must be a difference in the way the saves are executed. Wonder if the answer is there ?

[Martin Gühmann]

Originally posted by stankarp
When CTP 2 first came out it had a save game bug, and from memory, it probably started to happen about the same time (turns wise ) as cradle(300). The official patch 1.1 was primarily designed to fix that. Maybe it would help to look at the patch. It always happened about the time i got gunpowder and cavalry.

In cradle it happens about turn 300 onwards as well. I tried resetting maxplayer to 0 and restarting the game and same thing is happening again at about the same time. I tried changeing the capture city code to have a chance of capturing a tech when I capture a city (as per Martins suggestion) but this did not work as well. Although I restarted the game rather than start a new one.

I never noticed this bug I only noticed another savegame bug after the patch if I am using the scenario structure e.g. premade maps. The bug was an extra turn bug I saved a game reloaded it and all the movement points of my units were reset, my cities good extra production, I got extra gold , extra PW and so on. The interesting thing is that Cradle 1.1 used the scenario structure and I know that this bug didn't occured in a Cradle game, maybe it could lie in one of the text files, I only know that the default text files and the patched ctp2.exe without the text files from the patch causes also this exrtra turn bug.

Originally posted by stankarp
Interesting that it only happens when the player saves, not during auto save. There must be a difference in the way the saves are executed. Wonder if the answer is there ?

The difference between normal saves games on the one hand and quick saves and auto saves on the other hand is that quick and auto saves are smaller. In addition a normal save game contains an uncompressed *.rim picture with the powergraph and a map on it that can you see on the preview screen on the load window.

-Martin

[hexagonian]

Come to think of it, in my last game, I was using both an altered maxplayer setting and a modified map setting size (between the normal gigantic and and custom gigantic map setting that I had set up for Cradle) and I was getting the crashes too (this was the first time I had any type of ingame crashes since I got the game). Autosave helped somewhat (in that I had to end a turn before quitting the game to save the turn - then launching the Autosave file when returning to the game) and there were times that I could also normally save the game, but it was inconsistent.

If players could report all factors that they are using in their game when it crashes, then we can narrow down what is causing the problems. It is entirely possible that the altered maps sizes and the maxplayer settings may be changes that the game engine may not have been designed to support - thus the crashes.

Has anyone tried any of the new Victory options? I'm wondering how easy/hard they are.

[stankarp]

Original prep patch bug.
I remember it was a slightly different bug. The save files were corrupted and would not reload, causing the game to crash, rather than crashing during the game.

Autosave.
Autosave does not appear to properly record PW in progress. At least 2 times I have reloaded from autosave when cradle crashed and roads that were completed before the crash were still being built when reloaded from autosave. Also, I am sure that sometimes the turns to next advance has defaulted back to the start for that advance but I cant say for certain. Curious

Settings for Cradle crash.
Maxplayer 0 or 20
Capture city code normal or altered.
Ultra gigantic map.
15 civs.
Turns - about 300 onwards.
I am playing on by managing my turns, ie, only saving after the start of a turn and only when there appears to be nothing critical happening and preferably just after my last advance.

Questions. Do ZOC extend into mountain squares. I had a barbarian legion move through ZOC in mountain squares. At first I thought it was because one blocking unit was a horse archer who cant go into mountain squares but I used a legion next time and the same thing happened.

[stankarp]

I tried something new yesterday. I let the computer create its new turn based save file each time I saved, rather than oversave an existing pre named file. So far, 8 or 9 saves, and touch wood, NO CRASHES.

[Martin Gühmann]

Originally posted by stankarp
I tried something new yesterday. I let the computer create its new turn based save file each time I saved, rather than oversave an existing pre named file. So far, 8 or 9 saves, and touch wood, NO CRASHES.

That must be the reason why I never encountered this problem, since I got CTP1 I am used to it to create everytime a new save game and some time late I will delete them from windows explorer. This procedure was quite pritty neccessary if you don't want to lose half of your empire to the AI entity.

-Martin

[hexagonian]

Peter Triggs DiploMod Beta file now available

stankarp
Thanks for the update on file crashes - hopefully this procedure will help drcrease them

RE: ZOC
I'm stumped on this because both of the units mentioned do not have the NoZOC flag in them. Perhaps it is a terrain specific thing...

HERE IS AN UPDATE ALERT

Attached is the Cradle Beta version of Peter Triggs Diplo rewrite file - below is the email that he sent me that has the info about what he has done. I am really hoping that we can get some players to test this, so that it can be improved upon, and also because so many other Modders are taking a great deal of interest in this rewrite for their own Modwork. The files will not overwrite any existing Cradle files, but will create a new option called CRADLE 1.3 (Triggs Diplomod)

Peter Triggs Diplomod rewrite (Beta version)

A) If you look at NewDiplomod.slc, you'll see why it's taken me so long. I think it can safely be said that this is the most ambitious SLIC project yet undertaken. Somehow I got the idea that I should 'expose' a big chunk of the AI's response logic and this is what most of the code in there is doing. As a result it's virtually like we've got the source code in front of us. But this was a learning experience and if I were starting over from scratch I don't think I'd attempt to do it this way, it was just too time consuming.

I guess the main features of NewDiplomod are:

1) AI-AI Diplomacy is implemented.

2) Embassys are established by negotiation. (Actually, they come virtually automatically after the discovery of Dynasty. This might be too easy.)

3) No Trespassing, Ceasefire, Peace, Trade, Research, Military, and Alliance treaties all have fixed lengths.

B) I changed the trigger conditions for the surrender handler to those you suggested: it won't work until after 600 game turns and the AIciv has only 3 cities left. For the moment I've left the conditions for the messages unchanged: when you get a "We've won" message, that's the point at which they would have surrendered based on a military rank difference of 25. I think you'll find that it's really not that easy (with either setting) to get a civ to surrender, but I'm not sure that either of the above possibilities is the one we're after.

I've also left a few other messages in there that will spew out no end of data for your enjoyment. These will have to be taken out before releasing the thing on the public.

C) There's also a version of CRA_tileimp.txt where I used Sun_Tzu_159's technique to get the AI to terraform. If you compare it with the original you can see how it works and whether or not you like the specific chain of improvements I/he used. (Can't remember how much of his data I changed.)

D) While perusing some of the AI data files, I noticed that you had a BUILDING_BUILD_LIST_START but hadn't put it into any of the buildlist sequences.
Did you decide not to use it? There's a copy of CRA_buildlistsequences.txt where I've put it into all the sequences except the military ones. Maybe you've already tried this and didn't like it, but anyway it's there for you to have a look at.

E) I've included a revised version of the UnitUpdater that fixes the bug where if you try to update a unit in a stack of 12, you only end up with a stack of 11 and also the bug where you got the message about 0 units. Since these were the only bugs that have been reported since last Summer, I think it's safe to say it's bug free.

F) In CRA_script.slc, I've outcommented CRA_frenzy.slc. When I had a close look at it the other day, I was amazed at how many bugs I found. I doubt that it does much and what it does do is pretty much of a fluke. I suggest playing without it for awhile and seeing if you notice much difference.

Later comment: I don't know how that thing works, but it does do something under some circumstances. Playing without it for the past couple of days, I just don't see the big stacks.

G) Given the size of NewDiplomod, the chances that it's bug free are negligible. Also, I think it's going to need some balancing, but that's what beta testing is for.

Almost every time I've tried it I've got a different sort of game. I'm worried that it unbalances the science dimension in particular. See what you think, it is adjustable.

The file is below - as usual, please report any missing files, game crashes ASAP, although Peter has been testing these files on his Cradle setup.

This file has also been updated as per spanscape's post below

[Locutus]

Re: Peter Triggs DiploMod Beta file now available

Originally posted by hexagonian
I am really hoping that we can get some players to test this,

Woohoo! (we need a : party : smily) I've thrown my agenda and to-do list out the window. If anyone needs me: tough luck, 'cause I'll be playing Cradle all night
(Unfortunately I have to work all weekend so I can only play tonight )

[Spanscape]

Thought I'd give the new Diplomod a try, and I first got these path errors:

CRAG_govern.txt not found in asset tree.

To get past this, I pointed CRAT_gamefile.txt to CRA_govern.txt

Then, I received:

CRAG_buildings.txt not found in asset tree.

Got past this by pointing CRAT_gamefile.txt to CRA_buildings.txt


At long last I end up with some SLIC errors:

CRAT_NewDiplomod.slc:6637:M_TestBeginTurnAgreement s not found in string database.

CRAT_NewDiplomod.slc:6658:NDM_sender_science_resul tM not found in string database.

Couldn't find these anywhere - I figured they'd be in CRAT_NewDiplomod_str.txt?

Nice job though from what I see, damn near 7000 lines of SLIC in that one SLC file.

[hexagonian]

Spanscape,

I accidently picked up an old gamefile.txt file from my work computer and used the old file names that were in that file. Your fix is the way to fix this, but I am also attaching a new zip file below that corrects this problem, for anyone who wants the correct version.

I got the same SLIC error messages too - this will have to be addressed by Peter, but I am sure he will be keeping an eye on this thread for comments about his work.

One thing I did note, the game can be relaunched from the desktop after the crashes and played - at least I tried it for a few turns.


(Slightly) Fixed Beta file below...

[Immortal Wombat]

Re: Re: Peter Triggs DiploMod Beta file now available

Originally posted by Locutus
Woohoo! (we need a : party : smily)

Yes. Yes we do










6600 lines of SLIC, that is one immense script Peter, truly congratulations and thankyou for that huge undertaking.

[Peter Triggs]

Thanks, Spanscape and sorry about that. Looks like when I cleaned out the junkstrings from CRAT_NewDiplomod_str.txt, I left a couple of junk messages in the SLIC file. I've attached an ammended version with the messages removed.

A couple of points:

If you think that the AI civ's are not building enough units at the start of the game, go to CRAT_gamefile.txt and change CRAT_buildlistsequences.txt to CRA_buildlistsequences.txt; this will restore the original Cradle data.

As I mentioned in my e-mail to Dave, it's hard to see what FrenzyAI is doing but it must be doing something. To re-instate it, go to CRAT_script.slc and at the very bottom of the file remove the two forward slashes so that //#include "CRA_frenzy.slc" becomes #include "CRA_frenzy.slc"

I'm having another look at this code at the moment.

[Spanscape]

On another note, perhaps someone can help me get to the bottom of the multiplayer crashing issues I'm seeing...

When I play multiplayer (myself and my girlfriend) versus 6 AIs, the game will crash whenever we conquer an AI civ. This crash occurs whenever the conqueror of the civ ends their turn and it starts to cycle through the AI turns.

The crash occurs on both MedModII and Cradle. It does *not* happen with the ApolytonMod.

The crash occured tonight playing stock Cradle 1.3 on a *small* map with 2 human + 6 AI civs *very early* in the game. I have attached my savegame _right_ after I conquered the civ and before I ended the turn.

Thanks for any insight... I'm digging through SLIC, but its a new world to me. =)

[Spanscape]

Ok, interesting discovery:

I tried to load up the game I attached earlier, after I had conquered the 1st civ, and it would not let me launch it in multiplayer mode. I could copy it over and launch it single player mode without a hitch.

So I played it again from an earlier save and saved it _right_ before I conquered that 1st civ. That save game loads up fine in both multiplayer and singleplayer mode.

It looks like something bizarre happens when a civ is conquered in MP -- maybe it realigns the list of civs and replaces the human player with an AI player or something?

Anyway, game is attached. You can load this one up in multiplayer mode with 1 person, conquer the obvious city and watch it crash...