A superior converter

[King of Men]

Would there be any interest in writing a converter for CK to EU2 that does a better job than the ingame one? I could contribute some coding, once my move to California is over with in the beginning of October; or if someone else could start earlier, that would be splendid. For now, though, I'm just sounding out the board to see if there's any interest at all. I would also like to hear what people think should be done by the converter.


Latest version :

Source code and class files.

Version for converting DV games:

Source code and class files.


Several people are having difficulty running this, so here are some instructions :

Edit: Before trying it this way, wryun made a GUI wrapper which perhaps is a little more intuitive. I haven't tried this myself, so I make no guarantees about it working, but try it first anyway.

GUI wrapper

The file contains source code (something.java), class files (something.class), and data files (either something.txt or just a name, I wasn't really consistent with this). You will not need the source unless you intend to make changes to the program. Unless you know how to program in Java, this is not particularly recommended. And if you do, you know how to do things and don't really need my help. So, from now on I'm going to ignore the source. The class files run the program, as described below. The data files contain a lot of parameters; you can change a considerable amount of the converter's working by editing the data files. Be careful, though; the program is not robust, and can be downright unfriendly if you give it a data file in an un-expected format. The data files are described in the README file.

So. You will have unzipped the file somewhere; put the savegame you'd like to convert in the same directory. Now get a command window. (I assume, btw, that you are on a Windows box.) Go to the directory, and enter

java -cp . Converter <MySaveGame.eug>

Please note that java is case sensitive; you must enter 'Converter' with a capital C. If you are lucky, this will run the program, producing an output file called 'convertedMySaveGame.eug'. It is rather more likely that you will get 'java is not recognised as blahblahblah'. This means that your command window doesn't know about the Java Virtual Machine. There are two possible causes for this :

1. The JVM is not installed.
2. It is installed, but not included in your PATH variable.

We will deal with the second option first. Try looking through your Program Files directory, or running a search, for the java virtual machine. It should be an executable called simply 'java'. If you find one, put the location of that file in your PATH variable. You can find instructions on how to do this here, among other places.

If you can't find the JVM, you will need to download it. You can find it here. You will need to change your PATH variable to include its new location, as above.

Hope this is helpful. If it turns out not to be, tell me where you got stuck, along with the text of any error messages you get, and I'll do my best to extract you.

 

[unmerged(45754)]

Perhaps make it so that your rulers of your nation in EU2 are of your CK dynasty? I don't know if that's possible, but that'd be really kick ass.

Fixing what nations get what.. certain duchies seem to go independent no matter what.

Certain Muslim nations becoming Shia, such as the Fatimids

Show what actual lands you'd know. As it is in the game, even if you're playing an eastern nation, or say the Kingdom of Jerusalem, you only see what your stuff is, even if you are touching it, and were warring with it in CK.

Perhaps clear up some tags and such so tha they're a bit more realistic. King of al-Muratibids should != Mamuluks in EU2.

Perhaps some way to clear up cores, so that if you have a random claim for Scandanavia and you are based in Italy, you don't get a core, don't see a feasible way that this'd be done.

King of Germany not converting as Kingdom of Germany, but rather converting as new boundaries for the HRE as it is in game.

 

[Akka le Vil]

More than anything, fix the parameters of the kingdome. As it, we always start with "centralization" at max and others.

 

[unmerged(45754)]

Ah, yes, very good point! I didn't even think about that one.

Also, what decides where a CoT is? Because in my converted games, I get like 3 of them all next to each other, and then some areas that don't have them anywhere near.

 

[gigau]

Certain Muslim nations becoming Shia, such as the Fatimids

Yeah, and usually, in my converted games, each time Baden goes independant, it becomes Birma, with indhou-like troops. :rofl:

Also, what decides where a CoT is? Because in my converted games, I get like 3 of them all next to each other, and then some areas that don't have them anywhere near.

I agree that the CoT are sometimes weirdly placed... :wacko: CoT should be placed in the richest places of the map....

Another point... when you play CK, you have access to all map. But when the game is converted, some maps get hidden. I think, in the converted game, we should have access to the whole CK map (not more, but not less ).

 

[Grosshaus]

Don´t forget the AI files

 

[King of Men]

OK. Let me try to list the issues, and we can think about them a bit more systematically. If I miss anything, please help me out. I'll try for an approximate order of priority.

• Actual land borders. The builtin converter manages this pretty well, so it's kind of a minimum target to aim for! The main questions here are
  1. Should claims carry over into shields, or should these be hardcoded? I would be more inclined to leave claims as a timed CB, say of fifty years' duration, and give nations shield on the historical territories of their kingdom titles.
  2. Dukedoms. Vassals, incorporate them, or a mix? If a mix, how should it be decided?
• Slider settings. What should affect them, and how?
• COTs and province wealth. Suppose we say that each major region has a COT, and put it in the wealthiest province in the region? Maybe with bonuses for cities that historically were trading centers?
• Known areas to start with. I don't think knowing the entire CK map is the least bit unreasonable, that's what you start with in the vanilla EU2 GC, after all! But the ME powers should maybe know a bit of India and Caucasus also, and any Egyptian powers perhaps a bit of the Nile.
• Kings. It is not impossible, by any means, to generate random kings with your dynasty and cultural names - they're just text lists, after all. But do people want this? The EU2 versions don't make a lot of sense, in some cases.
• Events. I do feel that EU2 without events is slightly bland. On the other hand, a CK campaign is rather unlikely to generate anything remotely resembling the historical setup - the Ottomans won't even exist, short of being hardcoded! Now, it's possible to work around this by assigning CK powers to regions, and giving the major power in a region the events belonging to the historical power. For example, the most powerful kingdom in Italy gets the Venice events, that in Spain gets Castille, and so on. This does cause some difficulty in the Middle East, which is likely to be completely rearranged by successful crusades. A Kingdom of Jerusalem could get the Knights of Rhodes events, maybe? Seljuk Turks could stand in for Ottomans, and perhaps the Byz events could be borrowed from AGC. I would love to hear suggestions on this.
• AI. Again, perhaps we could just assign the files of approximately equal powers? This should actually be easier than the events, since the geographic strategies should be much the same whether the dominant power in Turkey is called Byzantium or Ottoman. But really, I expect this to be mainly used in MP games, so AI is maybe not that much of a concern.

 

[unmerged(21937)]

Regarding claims and core shields. I think that county claims should transfer as shields, but duchy or kingdom claims as mere 50-year CBs. County claims represent how your ruler considers the province as lawful possession, while claims on duke or kingdom titles don't really imply claims on any land area. Also you should get natural areas of all kingdom and duke titles your ruler holds as core shields. Not for counties however. So king of France, duke of Savoy and count of Genoa would have France and Savoy+Piemont as core provinces, but not Genoa.

Sliders, laws should be play very important role here.

For CoTs, I think it should be hardcoded and CoTs placed in important historical trade nexuses. Or maybe having it vary by wealth and / or ownership in a few province area.

For known areas, I think you should have a few sets of provinces and assign the sets by capital location.

I think a randomly generated king list is much better than having fixed set of monarchs that often make very little sense in a converted game. Randomly generated lists would also be more in the spirit of CK.

Events, I think an expanded, more interesting and bigger effect random set is best option. Again, having fixed events for specified nations is not in CK spirit and is boring.

AI, have few templates and assign them based on factors like power, sliders, being coastal or not.

Then, like kings, leaders should be randomly generated.

 

[Captain Frakas]

County claims represent how your ruler considers the province as lawful possession.

In EU2, core province also represent how the province population considers the ruler as lawful ruler.

 

[unmerged(45754)]

Out of all of them, I'd really have to say that the randomly generated list of leaders would be an absolutely excellent addition and probably the change most in the spirit of CK. I mean... CK is basically ALL about your dynasty, and if you have some random dynasty which is in no way related to your dynasty, what's the point, other than having conquered land, something that CK isn't necessarily about.

For example in my game as the Torchitorio's of Sardinia, I've essentially spread the Torchitorio name throughout all of Europe. Major nobles in France, Scandanavia, Britain, and even fairly major nobles in the remnants of Byzantium are all Torchitorios.

Also, perhaps you could speak to Grall or Grell about his mod that allows certain Kingdoms to pop up, ala mongols? That might like to a setup that could, in some ways be more historically, but wouldn't necessarily end up the same way.

As for duchies... perhaps it should be a variety of factors that decide whether it should be independent, part of the nation, or a vassal of the nation, such as loyalty, etc.

Also, perhaps you should open up discourse with the creators of the EU2 -> Victoria engine so that you don't create anything that would be represented oddly in Victoria. (As a lot of people who play games from CK -> EU2 also take them to Victoria, and perhaps HoI2. Though playing a game all the way from CK to HoI2 would be a VERY VERY different game, historically.)

 

[Enravota]

Maybe introduce some of the cultures not represented in CK (Slovaks, Romanians, Albanians etc)

 

[gigau]

[*] Actual land borders. The builtin converter manages this pretty well, so it's kind of a minimum target to aim for! The main questions here are

I agree... although there are some problems like Baden turning into Birma and Fatimids turning into Xhia.

Should claims carry over into shields, or should these be hardcoded? I would be more inclined to leave claims as a timed CB, say of fifty years' duration, and give nations shield on the historical territories of their kingdom titles.

I would think sheilds should be for a limited number of counties.... Once, in a converted game as France, i had shields all over Europe. As for duke and king claims, i agree with Biakhiam for 50-years CB.

Dukedoms. Vassals, incorporate them, or a mix? If a mix, how should it be decided?

IMHO, it should depend on loyalty.
If loyalty = 100%, incorporation
If loyalty between 99 and 10%, vassalization
If under 10%, total independance

Slider settings. What should affect them, and how?

Definitely should depend on the situation of the kingdom at end of CK. For example, if a monarch owns very few duke titles, and has lots of duke vassals, clearely, the EU2 converted kingdom should be decentralized.

COTs and province wealth. Suppose we say that each major region has a COT, and put it in the wealthiest province in the region? Maybe with bonuses for cities that historically were trading centers?

I have been thinking about that question for some time... I think that would be a good idea :1 CoT per region, in the richest province. I don't think a bonus to historical cities is necessary.

Known areas to start with. I don't think knowing the entire CK map is the least bit unreasonable, that's what you start with in the vanilla EU2 GC, after all! But the ME powers should maybe know a bit of India and Caucasus also, and any Egyptian powers perhaps a bit of the Nile.

That would be great.

Kings. It is not impossible, by any means, to generate random kings with your dynasty and cultural names - they're just text lists, after all. But do people want this? The EU2 versions don't make a lot of sense, in some cases.

Either way is good to me. Maybe a slight preference to randomly generated monarchs.

Events. I do feel that EU2 without events is slightly bland. On the other hand, a CK campaign is rather unlikely to generate anything remotely resembling the historical setup - the Ottomans won't even exist, short of being hardcoded! Now, it's possible to work around this by assigning CK powers to regions, and giving the major power in a region the events belonging to the historical power. For example, the most powerful kingdom in Italy gets the Venice events, that in Spain gets Castille, and so on. This does cause some difficulty in the Middle East, which is likely to be completely rearranged by successful crusades. A Kingdom of Jerusalem could get the Knights of Rhodes events, maybe? Seljuk Turks could stand in for Ottomans, and perhaps the Byz events could be borrowed from AGC. I would love to hear suggestions on this.

Tricky question...It depends on what we want. Do we want a longer CK game ? or do we want to make a change and play EU2 but with a CK-generated map ? As for me, when i convert a game, i want to play EU2. If you take the events from EU2, i think you take away a great deal of the game. Of course, i understand that if Portugal isn't created, the Tordesillas treaty can't be signed. But, in a EU2 game, i've already seen that treaty signed although Castilla had already swallown Portugal. I would prefer to keep the events from EU2.

AI. Again, perhaps we could just assign the files of approximately equal powers? This should actually be easier than the events, since the geographic strategies should be much the same whether the dominant power in Turkey is called Byzantium or Ottoman. But really, I expect this to be mainly used in MP games, so AI is maybe not that much of a concern.

No idea on that
Except on the AI colonization : Spain, Portugal, England,.... don't go on exploration and don't colonize if the converted file isn't updated. That shold be corrected. Maybe if Castilla doesn't exist at end of CK, give it's AI to Leon... And if England is dead, give it's AI to Ireland or Scotland. I know it's not quite accurate, but it's better for the gameplay.

Maybe introduce some of the cultures not represented in CK (Slovaks, Romanians, Albanians etc)

I agree, I would like Occitan not to be converted into italian culture. I'm from southern France. Occitan is a latin-type of culture, but it's occitan, not italian....

 

[unmerged(45754)]

Perhaps give colonization AIs to nations that are on a coast, and have certain sliders set a certain way?

Giving certain AIs to certain nations makes sense (Such as the Ottoman AI for the Byzantines if they are still fairly strong after the conversion.)

Or perhaps have it just randomly assign AIs? As in an ahistorical setup, there'd be no clue who was doing what, necessarily.

 

[unmerged(21937)]

In EU2, core province also represent how the province population considers the ruler as lawful ruler.

Not always. For example Ottoman lots of cores, Chambers of Reunion cores or Russian plethoras of cores don't afaik fit into that definition.

Gigau, do you perchance try to play converted scens in AGC-EEP? That's not really supported at all by the Paradox converter, since AGC-EEP folks have recycled plenty of CK converter EU2 tags into use elsewhere. There's no Birma or Xhia in vanilla EU2.

 

[gigau]

Not always. For example Ottoman lots of cores, Chambers of Reunion cores or Russian plethoras of cores don't afaik fit into that definition.

Gigau, do you perchance try to play converted scens in AGC-EEP? That's not really supported at all by the Paradox converter, since AGC-EEP folks have recycled plenty of CK converter EU2 tags into use elsewhere. There's no Birma or Xhia in vanilla EU2.

As a matter of fact, yes. I tried to play with AGC-EEP which brings more fun when playing EU2 (according to me at least ). Do you have a solution for me to solve that problem ?

 

[unmerged(21937)]

As a matter of fact, yes. I tried to play with AGC-EEP which brings more fun when playing EU2 (according to me at least ). Do you have a solution for me to solve that problem ?

Well, playing with vanilla EU2 immidiately solves the problem. Otherwise you will need to change AGC-EEP's country definitions or just part of them, into vanilla versions. All-in-all, isn't a big part of AGC-EEP's goodness in the scenario setup and in converter you will not have that?

Btw, King of Men, some kind of support to modded EU2 maps could be nice.

 

[King of Men]

Modded maps?! OK, I'm going to have to pull a MrT here and say "Send $$$$$$$$$". Sorry, but there are limits to the coding time I can devote to a hobby project.

Now, this is an excellent point :

Tricky question...It depends on what we want. Do we want a longer CK game ? or do we want to make a change and play EU2 but with a CK-generated map ? As for me, when i convert a game, i want to play EU2. If you take the events from EU2, i think you take away a great deal of the game. (...) I would prefer to keep the events from EU2.

In fact I think this is a fairly major design decision that should be taken as soon as possible. I myself am in favour of making a converted game be an EU2 game; if you want to continue playing CK, there is surely a no-time-limit mod out there somewhere. I'm willing to hear contrary arguments, though.

IMHO, it should depend on loyalty.
If loyalty = 100%, incorporation
If loyalty between 99 and 10%, vassalization
If under 10%, total independance

I don't think loyalty is a good indicator, it's too volatile. The average loyalty of the dukes might better go into Stability, which is the same kind of volatile indicator of short-term troubles. It's just not very difficult to ensure that all your dukes have 100% loyalty at the moment of conversion, though, so it might not be worth the bother.

I think distance from the capital, power, and relation to the dynasty would be better factors. Distant dukes are more independent; powerful dukes are more independent; dukes of a different family are more independent. As a general rule, though, I would prefer to have as many vassals as possible. With the exception of England, it's more historical - for that matter, even England still has a couple of crown duchies with privileges not given to other corporations - and I also think it makes for better gameplay. Challenge is good! It'll be more difficult to keep your kingdom together if it's a bunch of fractious vassals that somebody might annex.

Definitely should depend on the situation of the kingdom at end of CK. For example, if a monarch owns very few duke titles, and has lots of duke vassals, clearly, the EU2 converted kingdom should be decentralized.

I think the converted kingdom should be decentralised anyway; I don't think we should be seeing anyone come out with more than 5 Central, and 0 should not be unusual. As Byak says, the laws should be important. The size of the kingdom might be made a modifier; say, if the number of provinces exceeds some size, you get -1 to your Central. Or we could do it as a ratio of duke vassals to duke titles held by the king.

 

[unmerged(21937)]

Modded maps?! OK, I'm going to have to pull a MrT here and say "Send $$$$$$$$$". Sorry, but there are limits to the coding time I can devote to a hobby project.

It wouldn't need to be support for every existing custom map, but rather allowing the person using the converter have options to either "convert with standard map" or "manually redefine what each CK province corresponds to in custom EU2 map".

In fact I think this is a fairly major design decision that should be taken as soon as possible. I myself am in favour of making a converted game be an EU2 game; if you want to continue playing CK, there is surely a no-time-limit mod out there somewhere. I'm willing to hear contrary arguments, though.

I'll give you one. For me, the biggest drawbacks of EU2 are it's slavish devotion to history, which especially sucks in CK-converted game. If Habsburgs never got away from Aargau in CK, we are still trapped in Habsburg world in EU2.

To me the beauty of conversion is getting to play your converted game futher in EU2 world. The whole world, with it's exploration, colonization, trade etc. Now, events are all fine and dandy, but historical events are hopelessly tied to history and recreation of it. Just like monarchs and leaders.

Fortunately, events can be easily enabled / disabled in EU2, so a converter does not need to say anything about them really. And if I were you, I'd have options to either randomize monarch- and leadertables or use normal ones. Having both options available would be best choice, since it would please both audiences.

I don't think loyalty is a good indicator, it's too volatile. The average loyalty of the dukes might better go into Stability, which is the same kind of volatile indicator of short-term troubles. It's just not very difficult to ensure that all your dukes have 100% loyalty at the moment of conversion, though, so it might not be worth the bother.

I'll echo an agreement on that.

I think distance from the capital, power, and relation to the dynasty would be better factors. Distant dukes are more independent; powerful dukes are more independent; dukes of a different family are more independent. As a general rule, though, I would prefer to have as many vassals as possible. With the exception of England, it's more historical - for that matter, even England still has a couple of crown duchies with privileges not given to other corporations - and I also think it makes for better gameplay. Challenge is good! It'll be more difficult to keep your kingdom together if it's a bunch of fractious vassals that somebody might annex.

Of course available EU2 tags limit this one. For example in France or Germany there are enough tags to have a state in almost every province. On the other hand, in Scotland there in not a single tag to make a vassal of.

 

[gigau]

I don't think loyalty is a good indicator, it's too volatile. The average loyalty of the dukes might better go into Stability, which is the same kind of volatile indicator of short-term troubles. It's just not very difficult to ensure that all your dukes have 100% loyalty at the moment of conversion, though, so it might not be worth the bother.

Well thought

I think the converted kingdom should be decentralised anyway; I don't think we should be seeing anyone come out with more than 5 Central, and 0 should not be unusual. As Byak says, the laws should be important. The size of the kingdom might be made a modifier; say, if the number of provinces exceeds some size, you get -1 to your Central. Or we could do it as a ratio of duke vassals to duke titles held by the king.

On the other hand, if one has kept most of his duke titles, the kingdom can be considered as somewhat centralised...

 

[unmerged(21937)]

On the other hand, if one has kept most of his duke titles, the kingdom can be considered as somewhat centralised...

Not necessarily, if the kingdom has Feudal Contract law and huge lots of count vassals. I think the proportion of realm held directly in demesne vs held by vassals in CK should have significant effect in centralization. Of course it would mean that bigger kingdoms would by necessity be less centralized than smaller ones, but I'd think that would be realistic.