Kingdoms from the Scratch

[unmerged(27106)]

When I re-ran the scenario, I got the CTD in 1191. It could be a character's dynasty issue, where a character has a faulty dynasty code that triggers a crash when s/he dies- much like the bug that showed up in 1.03a. I'd go through and check it tonight, but my wife thinks I've bene on the computer too much already.

If I don't see anything posted about this when I check tomorrow, I'll take a peek myself.

 

[unmerged(21937)]

Although it seems odd when running the Solmyr's save from 1086 it always triggered on exact date, which should not be the case with deaths as they are not preset.

But I'll look into the originals then.

 

[Sterkarm]

Although it seems odd when running the Solmyr's save from 1086 it always triggered on exact date, which should not be the case with deaths as they are not preset.

But I'll look into the originals then.

Not necessarily, couldn't the death have been before the save?

 

[Solmyr]

I'll look at the scenario files and see if I can spot anything.

Might we worth rechecking that all dynasty entries on characters have type = 12.

Also, many starting characters have mothers but no fathers, could that cause problems for inheritance tracing?

If we locate the problem it might be possible to repair it in the save we have, so maybe all the posturing will still be valid.

Oh btw, as a pointless aside to tease Pat, in my test run George's son and Charles' daughter had a son with base stats of 10-10-8-10.

 

[Solmyr]

All dynasties seem to have the correct type, so that shouldn't pose a problem...

Anyone care to check that there are no duplicate or broken title entries?

 

[unmerged(21937)]

I can do a title check.

EDIT: Could a cause for CTD be that as initially royal demesne counties were vassals to duchy titles held by king and after king has given those duchy titles away, those counties held by king are still vassals to the duchy title.

Example: At start, counties of Friesland are vassals to duchy of Friesland. If king of Germany gives duchy of Friesland title to a vassal, his demesne provinces in Friesland will remain as vassals of Friesland in title section.

 

[Solmyr]

You know what, you might have something there. I noticed that two counties (C061 and C062) that were set a vassals of YORK but have been retained by ENGL still have YORK as their liege.

I'll tinker with this a bit and get back to you on whether it helps...

 

[unmerged(21937)]

Well, making all kings' personal demesne counties liege to be king should solve it if it's that?

 

[Solmyr]

It seems in regular saved games the liege of all counties held by a king is ----. Moreover, after the starting dukes die, all subsequent counties held by vassal dukes have the kingdom's tag as their liege. The latter however does not cause problems, but the former might. I'll try to figure it out...

 

[Solmyr]

Fixed lieges of all royal counties where needed, didn't seem to help, still CTD. I'm gonna fix the lieges in the original titles file and test-run them to check.

Also I found a bug in the original scenario titles file, the counties that should be held by the duke of Toledo are held by the king of Castile instead. That shouldn't be causing problems in the save though, as the duke's title simply disappears once the game starts and the king creates it for himself.

 

[unmerged(21937)]

My GF has asked me to come with her to her cousin's wedding tomorrow and I can't be sure that I'll be back by sunday morning. Maybe we should wait until next weekend with Kingdoms from the Scratch as it would also allow us to find out that CTD problem before game time for sure.

As the players in Game of Thrones are same as with Kingdoms from the Scratch, you (and I if I make it home in time) could play play that instead?

 

[Solmyr]

Here is the scenario with the royal county liege tags and the duke of Toledo bug fixed:

http://www.kolumbus.fi/solmyr/CK/KFS.zip

Running this, I got almost one CTD per year starting in the late 1080s and finally a repeatable one in February 1090. So I guess the county lieges aren't the problem here.

I'll start looking over the characters to see if anything looks wrong. Also, someone could check that country and titles files agree with each other about who rules what. Maybe some sort of check on the provinces file would also be good. Other than that I don't know what could be causing this.

I'm fine with playing GOT this weekend. It's okay if you aren't there Byak, we'll deal with Poland anyhow.

 

[unmerged(27106)]

I'll go through the country/title files here in a bit and see if I can spot any incongruities.

 

[Sterkarm]

Here is the scenario with the royal county liege tags and the duke of Toledo bug fixed:

http://www.kolumbus.fi/solmyr/CK/KFS.zip

Running this, I got almost one CTD per year starting in the late 1080s and finally a repeatable one in February 1090. So I guess the county lieges aren't the problem here.

I'll start looking over the characters to see if anything looks wrong. Also, someone could check that country and titles files agree with each other about who rules what. Maybe some sort of check on the provinces file would also be good. Other than that I don't know what could be causing this.

I'm fine with playing GOT this weekend. It's okay if you aren't there Byak, we'll deal with Poland anyhow.

Well, I think maybe instead of GOT, we could just play a small game, like maybe as KoJ in 1187 scenario. But, it seems counties/lieges is the problem actually, but maybe it's just being repeated after a while. Because, as you see, it didn't repeat on the same day as before (May 24, 1088)...

 

[unmerged(27106)]

Well, I think maybe instead of GOT, we could just play a small game, like maybe as KoJ in 1187 scenario. But, it seems counties/lieges is the problem actually, but maybe it's just being repeated after a while. Because, as you see, it didn't repeat on the same day as before (May 24, 1088)...

Why wouldn't we play GoT? I was under the impression that we were picking up KFS, but that we had agreed to return to GoT at some point in the future. I really don't have any interest in a one-off game when we could return to a scenario that we have put a lot of time into.

 

[Solmyr]

I found some bugs in the characters file, though none of them CTD-causing:

- duke of Deheubarth is Irish, should be Welsh
- wife of the duke of Salisbury is Catalan, should be English (unless Pat has some background on this that I'm unaware of)
- Edward Rose should be a brother of George (right?)
- wives of the dukes of Murcia and Catalonia are in the court of Aragon, should be moved to their husbands' courts
- wife of the duke of Silesia is in Sandomiersk
- country tag of char 1058 should be STEI instead of STIE
- country tag of char 1125 should be WIEL instead of WEIL
- Luca Hannus (id 1328) has Gabriel (1301) as her mother, should be father

I'll fix these and post the scenario again.

Additionally there are some things that may not be bugs but should be looked at IMO:

- Badajoz should be Castillan instead of Portuguese as it is capital of Leon
- provinces of Cyprus are Muslim Turkish, should be Orthodox Greek?
- Polotsk is capital of Lithuania but is Russian, should be Lithuanian?

The above is mainly so that we don't get unwanted culture changes for kings and dukes

I didn't find any other bugs in the characters file, will recheck again but we may have to look elsewhere for the cause...

 

[Solmyr]

Guys, I found the bug.

In the countries file, province 234 (Pavia) was both under ITAL and under MILA. However, ITAL should have had province 324 (Corsica) instead. As it was, Italy held the title to Corsica but did not control it. So in the late 1080s-early 1090, the AI king of Italy tries to give Corsica to his second son, and the game CTDs.

I've fixed this bug and the ones I pointed out above in the base scenario files, I've left the province culture changes above for discussion for now. I've also added an id meter at the beginning of the scenario to make it assign new ids only from 5000 upwards (in order to avoid id overlap with starting characters, just to be safe).

I've also fixed the Corsica bug in the save, and confirmed that it passes the CTD point just fine (AI Cristoforo indeed gives Corsica to his second son on the CTD date). However, I suggest that we restart the scenario from the beginning to incorporate all the other bugfixes (and maybe the province culture changes) and so that everyone can play their country in the beginning of the game when it's most critical. What do you guys think?

 

[Solmyr]

Also, if we do restart, I'd like to suggest that Germany's power is slightly reduced, if we want to avoid Game of Thrones 2 where everyone trembles in fear of German power. What I propose is that we move the duchies of Istria and Krain under Croatia, create Pomerania under Poland, and either move Brandenburg to Poland or Holstein to Denmark (or both). That would reduce the number of German dukes (not counting the royal duchies) from 19 to 16, still the most in the game but a bit closer to France's 14. If Germany loses both Brandenburg and Holstein then I'm willing to give up Bordeaux to Navarra as well. What do you think?

 

[Solmyr]

Downloads:

Fixed save from our last session (Corsica bug and royal county liege tags):

http://www.kolumbus.fi/solmyr/CK/Kingdoms.zip

Base scenario with all above bugs fixed:

http://www.kolumbus.fi/solmyr/CK/KFS.zip

 

[Hyzhenhok]

Great, Solmyr. I see how that would create the problem - we might have figured it out faster if someone had tried to get to 1100 playing as Italy.

I'll support restarting if Jarkko and Arco prefer and the others agree, though it certainly isn't necessary. On weakening Germany, I'm not sure it's necessary - the main reason in Game of Thrones that Germany became the monster is because suddenly there was no France to counter it. Maybe giving Holstein to Denmark would be good in slightly dumbing down Germany without doing too much. I'm strongly against giving Brandenburg to Poland; Poland is already stronger than Hungary.