Localized versions of Civ3

Quintillus

Resident Medieval Monk
Super Moderator
Supporter
Joined
Mar 17, 2007
Messages
9,306
What does everyone here know about localized (non-English) versions of Civ3? In particular, how to get English versions of Civilization III to display non-Roman characters correctly? Civ3's support for non-Roman characters appears to be virtually non-existent, but there's are active modding communities in Russia and China, so it would seem that there is a way to get at least some of these scenarios working with English versions of Civ.

As I recall, there are also more minor differences in display of fonts between the French and English versions. Do any of our French CivIII fanatics have both the French and English versions and experience with the differences? It wouldn't surprise me if other languages with Roman fonts, but with characters not commonly used in English, also have slight differences.

So far, what I've been able to figure out is that Russian uses the Windows-1251 charset, and I've added support for opening Russian BIQ files (with proper display of fonts) to the in-development version of my editor (though I need to test saving). But that's not as useful as getting it to display correctly in-game.
 
I would appreciate an answer for this too, as I intend to put city names in my Western World mod using the local name of that country, as well as names for use in the Civilopedia.

As far as I know, to get it to work with the Civilopedia, you would need to change the .txt file to use UNICODE and not ASCII. However, the game seems to crash if the civpedia is anything but ASCII. So I have no idea how it would work unless you could fix it.
 
To get around this, I use the local names, then romanize the characters. It is not the same I know, but better than using the English names...
 
Quintillus, I use RU version of the Civ3.
All Civilopedia texts written in Russian EXCEPT tags (for example, BLDG_Palace & DESC_BLDG_Palace remain in EN).
PediaIcons completely in EN.
Editor seems to be depended to the OS setting - I used Win-1251 earlier and it worked fine, now I use KOI-8R and it works fine as well. In editor I placed city names, resources/terrain names, rulers names, units names, worker jobs in RU - and it's correct.
And - if I modded under the KOI-8R and then started the game under the Win-1251, all names from my mod look like "??????". It's the known and understandable issue - so, to avoid it I think I should translate my scenario into EN, no problems with it anywhere. :)
 
I have the French version but I am currently using the English Pedia text for my mod. I'm not sure what to look for or what I can do to help you answer your questions, but if there's anything you want me to try or check, tell me and I'll do it :)
 
I have the french version too.
A utility called 'civpax' allow english mods to work under french version.
The name of the directories for the units is specific in the french version. For example you have a directory called 'Artillerie' instead of 'artillery'.
I think french version has a specific LSANS.ttf file
 
The name of the directories for the units is specific in the french version. For example you have a directory called 'Artillerie' instead of 'artillery'.
This is true for vanilla, not for Conquests: in the Conquests\Art\Units folder you will notice names like "Three Man Chariot" or "Tow Infantry" etc, proving that they were not translated in French (only the biq was translated).
 
To get around this, I use the local names, then romanize the characters. It is not the same I know, but better than using the English names...

I agree, and it also doesn't really help with playing mods that aren't designed for English Civ3 to begin with. Which is unfortunate - there are some interesting-looking mods for non-English versions of Civ, and it'd be a nice bonus that I'd pick up some Russian/French while playing them.

Here is an old thread:
http://forums.civfanatics.com/showthread.php?t=14666

Maybe the LSANS.ttf may be able to be modified to do as such. Seems the guy above may have done something like that, claiming to have localized all files.

I saw that thread while searching before creating this thread, but all the links are dead. Looked promising, but it's hard to learn more about it without any of the links (and even if they had worked, I'm not sure if it would've helped since my Vanilla is already fully patched and I'm not sure if I have the disc available to re-install the older-patch version).

Quintillus, I use RU version of the Civ3.
All Civilopedia texts written in Russian EXCEPT tags (for example, BLDG_Palace & DESC_BLDG_Palace remain in EN).
PediaIcons completely in EN.
Editor seems to be depended to the OS setting - I used Win-1251 earlier and it worked fine, now I use KOI-8R and it works fine as well. In editor I placed city names, resources/terrain names, rulers names, units names, worker jobs in RU - and it's correct.
And - if I modded under the KOI-8R and then started the game under the Win-1251, all names from my mod look like "??????". It's the known and understandable issue - so, to avoid it I think I should translate my scenario into EN, no problems with it anywhere. :)

Interesting. How do you go between Win-1251 and KOI-8R for the editor? :hmm: I wonder if setting my Language for Non-Unicode Programs setting in Regional and Language Settings to Russian will make Civ3 work better with Russian scenarios? I'll have to try. Although I still don't know whether it would default to Win-1251 or KOI-8R, as they're both checked in the Code Page Conversion Tables area. Edit: This did not work. Some programs did pick up the change, such as the clock which now reports the time in localized Russian format and language, and I can save Russian text in Notepad properly now, but Civ3 didn't pick up the change.

Are you able to play scenarios created by people with the English version of the game without difficulty? I have the issue with names looking similar to "??????" when trying to play mods that were created with the Russian or Chinese versions, even though if I set my editor to use Win-1251 and load the Russian one I'm testing with up in it, it displays properly.

I have the French version but I am currently using the English Pedia text for my mod. I'm not sure what to look for or what I can do to help you answer your questions, but if there's anything you want me to try or check, tell me and I'll do it :)

How do you use the English Pedia text with the French version? Or do you mean that your Civilopedia entries are written in English? Do scenarios written by people with the English version display properly for you?

I have the french version too.
A utility called 'civpax' allow english mods to work under french version.
The name of the directories for the units is specific in the french version. For example you have a directory called 'Artillerie' instead of 'artillery'.
I think french version has a specific LSANS.ttf file

Interesting. Is that utility somewhere on civpax.com? I see they have some translated scenarios there, including Escape From Zombie Island 2. They wrote there (original, and translated from the French for those who don't speak French):

Original French said:
Escape from Zombie Island 2 est exécutable sur sur Conquests en Français. Il s'agit d'un scénario initialement fait pour la version originale (anglais). Je le trouve super alors j'ai modifié les fichiers pour que les possesseurs de la VF puissent y jouer (c'était pas trop dur non plus). Les commandes et textes restent en anglais mais bon je pense que ce n'est pas trop grave, l'important est de pouvoir jouer. Il s'agit de la toute première version alors n'hésitez pas à me signaler tout éventuel problème durant le jeu (outre le fait que les textes soient en anglais). D'ailleurs si quelqu'un avait le courage de traduire les textes en français, qu'il me fasse signe ;). Une exclusivité civpax.com
Auteur: KingArthur / geo

Translated to English said:
Escape from Zombie Island 2 can be run on the French version of Conquests. It was initially made for the original version (English). I found it to be a super scenario, so I modified the files so that owners of the French version could play it (it wasn't very difficult either). The interface and text remain in English unfortunately, but I don't think that's too bad, the important part is to be able to play. This is the first version, therefore don't hesitate to contact me about any problems you encounter while playing (other than the fact that the text is in English). Finally, if someone has the courage to translate the texts into French, consider me impressed :). An exclusive of civpax.com.
Author: KingArthur/geo

Feel free to suggest improvements to the translation. My French is out of practice, but I figured I could still do better than Google.
 
Warning: I do not recommend running the fixpack at this time. It essentially changes your Civ3 to the Russian version. Which itself is fine. The issue is if you ever decide you want to go back to English. I haven't figured out a way to do that yet. The good news is, Civ3 in Russian should play essentially all English scenarios, too, with only a very few having minor visual issues.

Update (Feb 2014):: I have now figured out how to switch your font to English should you run the Russian fixpack. I plan to post more about this when I'm less tired (and ideally also post a language-swapper), but if I forget to and you want to know how, PM me.

Well, I've made some progress. I happened across this long-neglected thread. Figuring "what could go wrong with a download from a site with civfanatics in its name?", I downloaded and ran the fixpack.exe, after first making backups of all files with names related to LSANS, as well as the Conquests executable. After running the fixpack and restarting (per the instructions - I don't know if it is really necessary or not), I started Civ3. And, what do you know...

attachment.php


Civ loads up with Cyrillic support as well as English! Note that this is a Windows-1251 scenario. Also noteworthy is that the font is slightly different - a similar style, but not identical.

What did the fixpack change? The LSANS.TTF file is different - 113 KB versus 59 KB (or 49 KB? I had two versions in my folder, probably from past efforts to customize the font, and I'm not sure which is really the original). LSANS.FOT appears to have been modified as well, and it looks like it went ahead and converted my backup .FOT file as well (though not my backup .ttf files), so it's probably best to back up to a different folder. There also a few files that appear to have been left over from the fixpack running. At first glance it doesn't look like the Civ3 executable was modified, and the Firaxis editor still lacks Russian support. I'll be able to look into what was modified more carefully at a later time.

Score one point for general-jcl for being correct about LSANS.TTF being a solution!

This also only makes things work for Russian - Chinese support is notably still lacking, and I'd have to test other languages. It looks like a plausible solution would be adding support for more character sets to LSANS.ttf. But doing that is probably somewhat challenging.

Also perhaps relevant is that the executable reports itself as supporting Unicode. So my guess is the limiting factor is that the font that Civ3 uses only supports the characters that Civ3 shipped with in a given locale. The BIQ's text is locale-specific as well, but given that Civ3 can support both English and Russian scenarios at once, it appears to be able to translate them properly (though I'm curious how it decides whether a scenario should use the Roman or Cyrillic characters. Looking at the BIQ headers in the hex editor, I don't see any flags that might indicate which should be used, and all the undocumented ones in the GAME section match the most basic BIQ Firaxis's editor can create).
 
1 more interesting thing is that in Aeon mod (available in the DB) used a specific char, the christian cross. I couldn't insert it manually, but I was able to copy/paste it into my own .biq.

Next thought was "since there are religious building for other religions and there are characters for them in the Windows font table, then why don't use them?" I typed islamic crescent in the MS Word and copy/pasted - .biq determined as unknown symbol (empty rectangle, kinda []). OK, don't work that way - I have one more: I opened .biq with text editor (Akella), found a building name and copy/pasted there. Result was the same - [].

So, it'd be interesting to determine what causes correct usage of specific chars in the .biq - for example, CCM & Worldwide use religion-specific buildings as well and their authors may be interested in this finding.
 
How do you use the English Pedia text with the French version? Or do you mean that your Civilopedia entries are written in English? Do scenarios written by people with the English version display properly for you?

I've just downloaded the English pediaicons.txt and civilopedia.txt and dropped them into my mod's text folder. I've also made English-named copies of the folders that are in French in my vanilla folder so that I could play any scenario written for the English version.

Feel free to suggest improvements to the translation. My French is out of practice, but I figured I could still do better than Google.

You did a very good job, although I don't think "qu'il ma fasse signe" can be translated to "consider me impressed". I'm not sure how to translate this idiomatic expression but it would be something like "wave at me" or "call me" :)
 
cixpax is only a utility which automatize that:
" I've also made English-named copies of the folders that are in French in my vanilla folder so that I could play any scenario written for the English version."
...and i think we can still find it on civpax.com
 
1 more interesting thing is that in Aeon mod (available in the DB) used a specific char, the christian cross. I couldn't insert it manually, but I was able to copy/paste it into my own .biq.

Next thought was "since there are religious building for other religions and there are characters for them in the Windows font table, then why don't use them?" I typed islamic crescent in the MS Word and copy/pasted - .biq determined as unknown symbol (empty rectangle, kinda []). OK, don't work that way - I have one more: I opened .biq with text editor (Akella), found a building name and copy/pasted there. Result was the same - [].

So, it'd be interesting to determine what causes correct usage of specific chars in the .biq - for example, CCM & Worldwide use religion-specific buildings as well and their authors may be interested in this finding.

That looks like an interesting scenario - I had missed it before. I'm downloading it now (though sadly the Megaupload upload with some of the cumulative updates is missing). But the BIQs were quick to download, so I took a look at them. The cross character is 0x86 (base 16), which corresponds to the cross symbol in Civ's default Windows-1252 character set, wheras the dot is 0x95. I think this Wikipedia article will give you the characters that will work in English-language BIQs. And it looks like the cross and double cross are the only common religious symbols in that set. But if you wanted, say, a Yen symbol (¥), I think it probably would work.

So are you able to play Russian and English scenarios interchangeably? It would be nice if that came by default in the Russian version. By chance are you also able to play scenarios in Chinese with the characters displaying correctly? Edit: I just realized that the most common Roman characters are in the Windows-1251 character set as well, when I opened a scenario when my editor was still in RU/Windows-1251 mode. So that means that with the Russian version it probably does work just fine to play scenarios written for English, just not vice versa. It's only with a few of the symbols in Aeon that there were differences when I was using Windows-1251 mode, such as the AE combined symbol being replaced by the letter Ж, and most English mods just use standard Roman characters.

As a side note, it's currently required to use Firaxis's or Steph's editor for files with certain symbols such as the cross, as I incorrectly thought Civ was using the ISO-8859-1 charset instead of Windows-1252 before yesterday.

I've just downloaded the English pediaicons.txt and civilopedia.txt and dropped them into my mod's text folder. I've also made English-named copies of the folders that are in French in my vanilla folder so that I could play any scenario written for the English version.

You did a very good job, although I don't think "qu'il ma fasse signe" can be translated to "consider me impressed". I'm not sure how to translate this idiomatic expression but it would be something like "wave at me" or "call me" :)

Makes sense on the first paragraph. And thanks. "qui'il ma fasse signe" was one of the phrases I didn't really know, so I've learned something!

cixpax is only a utility which automatize that:
" I've also made English-named copies of the folders that are in French in my vanilla folder so that I could play any scenario written for the English version."
...and i think we can still find it on civpax.com

That makes sense!
 
No problem! I've also discovered, now that the Aeon mod has finished downloading, that the fixpack I mentioned in post 10 doesn't exactly add support for Russian (Windows-1251) to Conquests, but instead essentially converts the English Civ3 to the Russian version, functionally. As in, my Civ3 is now using the Windows-1251 character set. The text files are still in English, but since Windows-1252 (English) and Windows-1251 (Russian) both have the common Roman characters in the same places, the game reads them fine. It's only when looking at some of the symbols in Aeon that I could notice the difference. This also means that the game isn't detecting whether the BIQ is in English or Russian at all - it's just reading them with whatever character set its font supports, and it's a happy coincidence that the Russian version of Civ3 happens to support most English scenarios out of the box.
 
:bump: This has always been of interest to me. The ñ character is always displayed correctly, a friend of mine had it installed in Spanish and we used to play hotseat with him.

The characters for long o in Japanese appear to be a problem -I can't write them in the editor.
 
Is there a Japanese version of Civ3? It seems like each different locale has its own version (Chinese/Russian/etc. - I think most of western Europe probably is the same version, at least in terms of the characters it will recognize, even if the text was translated). I don't know if there's a Japanese version or not; if so its editor would probably be able to handle that.

I'm still not sure what exactly controls which locale's text the game can handle. I'm pretty sure it's not the executable, though. My copy's still stuck in the Russian locale, no matter which executable I use. That means that most English text works okay, but some things (usually accents and such) are off. But I guess I'd look and see if there is a Japanese CivIII fansite somewhere?
 
Update (Feb 2014):: I have now figured out how to switch your font to English should you run the Russian fixpack. I plan to post more about this when I'm less tired (and ideally also post a language-swapper), but if I forget to and you want to know how, PM me.

I obviously forgot about posting more on this after adding this to my post above, but I came across this thread again today and thought correctly that I had written more down somewhere. Sure enough, I found a file detailing what I did on my hard drive - what would I do without file timestamps?

So, the way I found to switch the font in-game involved registry editing. In particular, the following key is relevant:

HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows NT\CurrentVersion\FontSubstitutes

By default, it will contain a number of font mappings. On a fresh Windows XP install, these are:

Code:
"Arial CE,238"="Arial,238"
"Arial CYR,204"="Arial,204"
"Arial Greek,161"="Arial,161"
"Arial TUR,162"="Arial,162"
"Courier New CE,238"="Courier New,238"
"Courier New CYR,204"="Courier New,204"
"Courier New Greek,161"="Courier New,161"
"Courier New TUR,162"="Courier New,162"
"Helv"="MS Sans Serif"
"Helvetica"="Arial"
"MS Shell Dlg 2"="Tahoma"
"Times"="Times New Roman"
"Times New Roman CE,238"="Times New Roman,238"
"Times New Roman CYR,204"="Times New Roman,204"
"Times New Roman Greek,161"="Times New Roman,161"
"Times New Roman TUR,162"="Times New Roman,162"
"Tms Rmn"="MS Serif"
"Arial Baltic,186"="Arial,186"
"Courier New Baltic,186"="Courier New,186"
"Times New Roman Baltic,186"="Times New Roman,186"
"MS Shell Dlg"="Microsoft Sans Serif"

I believe this is what WordPad and Notepad use when you have options such as Western, Turkish, Greek, Cyrillic, and Baltic scripts. At any rate, when I added the Russian patch, it added the following entries:

Code:
"Lucida Sans,0"="Lucida Sans,204"
"Tahoma,0"="Tahoma,204"
"Courier,0"="Courier,204"
"Courier New,0"="Courier New,204"
"Arial,0"="Arial,204"
"Times New Roman,0"="Times New Roman,204"

After I removed the Lucida Sans line and rebooted, my Russian-language Civ3 was back to English. And adding just that line to a freshly-installed Civ3 on a fresh XP install flipped it to Russian, even without adding the modified LSANS.ttf file that the Russian fixpack added (although I am not 100% sure that all characters would display correctly; the most common Cyrillic characters did).

What does the 204 mean? It's a font character set, or font charset for short. This is not the same as a code page (aka character encoding); for Russian, the typical character set is 204, whereas the typical code page is Windows-1251. From what I understand they refer to the same thing, but font charset fits in the range 0 - 255, or a single byte in technical terms.

There's a list of font charsets at this MSDN page. It only goes up through Windows 95's support, but does contain charsets for some other languages as well, though not Japanese. I left Microsoft feedback suggesting that the documentation be expanded to include more recent versions of Windows; maybe they'll add it :mischief:.

I believe I also played around with this with Chinese so that I was able to play a Chinese Civ3 scenario in Chinese. Based on my notes, Chinese would be GB2312 in this table, and thus font charset 134, although there's the caveat that my abilities to determine if Chinese is correct are much lower than Russian, which is itself limited. I think I may have also had to map Lucida Sans to a different font for Chinese (perhaps SimSun?), although it's been long enough that I don't remember the details of that.

At any rate, hopefully this is of some benefit to those interested in playing scenarios in languages other than their native language.
 
Back
Top Bottom