Internationalization

[Khiraly]

I have committed a fix for this to CVS.

Hi,

I was recently examining the language .cfg files and discovered that certain text lines are not being stored the most appropriate way in my opinion. I am referring to the relation of the files sample_translation.cfg, english.cfg and the individual language .cfg-s. Those lines that should be translated should be stored in one file whereas now both the sample_translation.cfg and the english.cfg files contain some of them. This situation makes checking the difference between CVSs quite bothersome.
I would suggest creating one file for the keyboard shortcuts that are presently being stored in english.cfg while all the text could be stored in one file that should be used for translation purposes only, while keeping individual files for each language.

-------
There is a game feature issue I would like to recommend. Due to some playtesting , I have discovered that the use of fog of war could seriously improve the strategic possibilities and also pose some new challenges and decisions. There would be another use for quick, light units as they could make excellent spotters and it could emphasize more the differences between terrain types. Are you considering the idea?

Regards,
Khiraly

[Dave]

Khiraly,

sample_translation.cfg is generated from a tool called make_translation which is distributed with the game. Enter,

Code: Select all

make make_translation
./make_translation > data/sample_translation.cfg

To generate yourself a new, updated sample translation file. This will then contain all the strings from the English translation plus more that are needed, so you should never have to refer to the english translation directly, only to sample_translation.cfg

Once you have done a translation into Hungarian, and receive a new version of the game, and there are some more strings which need translating, you can easily see what these strings are,

Code: Select all

make merge_translations
./merge_translations data/translations/hungarian.cfg data/sample_translation.cfg > new-hungarian.cfg

This will create a file, 'new-hungarian.cfg', which contains all the already-translated strings in Hungarian, and the new strings which need to be translated at the top of the file.

Hopefully these tools should make translating easy for you. Let me know if you have any problems with them.

---

We are planning to implement fog of war as an option at some point, although it will probably not be implemented in the main campaign much or at all. If you have played to the end of the current campaign, you will have seen that some scenarios do use 'shroud'.

David

[Khiraly]

Code: Select all

make make_translation
./make_translation > data/sample_translation.cfg

Sample question: why it's needed to generate the english-string file?
Why it's not hardcoded in a something.cfg?
(I'm familiar with the gnu diff tool...)

Khiraly

[Dave]

Khiraly,

I'm afraid I don't understand your question. Can you rephrase?

David

[Dave]

There has been a major bug with Internationalization support in the past that accented characters at the beginning or end of a string would tend to get chopped off, and not displayed.

This bug has just been fixed, and a change committed.

Special thanks to Skovbaer for helping me track down this bug.

David

[Khiraly]

I'm afraid I don't understand your question. Can you rephrase?

Why it is necessary to create the proper sample_translation.cfg with a separate tool? I see no reason why it is not simply edited by hand. As I am quite used to the use of 'diff', it would be far more comfortable (for me, at least).

(I will not flame, it's just a question.)

Khiraly

[Dave]

Khiraly,

The reason why sample_translation.cfg needs to be generated by the use of make_translation is that it contains far more than just that in english.cfg. It actually walks through the entire data/ sub-directory, and finds all possible strings to be translated, including ones in scenarios for instance, which are not included in english.cfg.

It would not be easily possible to generate this by hand.

David

[DaringTremayne]

I downloaded the binary version for Windows of the CVS from 20/10/2003.
I was a little bit surprised to discover that some words or sentence that I've translated a month ago has disappeared in the French translation! (for example the skip button, or next button).

I've also discover that some sentences have been rewritten with mistake sometimes. I can admit that I also did some mistake and that in French an Elf could be called an "Elfe" (but I already saw the two ways of writting it), but for an Orc which is a totally unreal creature, we can't call them "Orque".
"Orque" are those big black & white whale.

Guillaume

Well I am sorry to have been at the origin of this misunderstanding ! Thats's quite involuntary, I assure you.

Now to address some of the points raised in this and other later posts :

1) I did not erase anything from the translation, but the file I used as the basis for my translation was the one provided with the game (not, as it appeared later, the most up-to-date). As my translation was primarily done for the benefit of my daughter, that didn't seem too important. Wen I later submitted my file, I was told it would be entered into CVS, which meant my changes would be undoable if they were not agreed upon by others.

2) as for my so-called mistakes in the translation, please note that I am not trying to impose my view of what is right. In particular, it is true that "Elfs" can be translated several ways, as well as "Orcs". I base my translations on the works of Tolkien, as translated by Ledoux in French, and as discussed at length in several forums. I can refer the French-speaking audience for instance to http://www.tolkiendil.com/ardalie/orc-orque.php (non-French speakers would probably find it boring anyway. Well, come to that, French-speakers might, too ). This does not close the debate, but hopefully will avoid hurting my pride by not having to talk about "mistakes" (I'm mostly kidding here really).

3) hopefully with the new organization we can work better together. However one point is not clear to me : where is the authorative "current translation" : in CVS or with the French maintainer ?

Hope this post clarifies things. I'm willing to submit my translations from time to time, this being such a very fine game, and I hope it helps.

Best regards,
Laurent (a.k.a. Daring Tremayne)

[Dave]

Well I am sorry to have been at the origin of this misunderstanding ! Thats's quite involuntary, I assure you.

Yes it is. I take full responsibility for causing these problems and apologize.

I'm willing to submit my translations from time to time, this being such a very fine game, and I hope it helps.

Yes, it would help, but could you discuss any proposed changes to the translation in the Internationalization forum, or send Guillaume (the official maintainer of the French translation) a private message?

That way we can have translators collaborating and discussing changes rather than overwriting each other's changes by mistake.

Thanks!

David

[DaringTremayne]

Yes, it would help, but could you discuss any proposed changes to the translation in the Internationalization forum, or send Guillaume (the official maintainer of the French translation) a private message?

That way we can have translators collaborating and discussing changes rather than overwriting each other's changes by mistake.

Thanks!

David

Sure, I'll do that. I'm planning to send my proposed translation to Guillaume, and I guess we'll discuss the most important issues on this forum. Well, the International one.

[DaringTremayne]

I don't know much about character encodings, and I don't know what's wrong with this file. However I have put it up at,

http://wesnoth.whitevine.net/files/polish.cfg

Perhaps someone else can look at it and tell us what's wrong?

David

Has this "problem" been settled already ? If so, disregard the following.

I believe the encoding is "ISO-8859-2". To check that, I have made a PDF file with some of the messages in the above file (not the ones in English ), which to my untrained eyes looks Polish... I could send this file to someone fluent in that language.

If that's correct, I wouldn't know what to do with it in the context of BfW ! but I htink we could work together to answer that, and I'm ready to do some experimenting (whit charset translators, etc.)

I'm waiting for your comments !

[Dave]

I don't think the problem has been solved....but I'm not sure

All I can tell you is that the only character sets I can support in-game are UTF-8 and LATIN1.

David

[methinks]

I know nothig about solving this problem, he. Anyway I've wanted to submit Polish translation updated today but I believe we will have to wait 2 more days. Sorry for this but I have some issues to solve in my real life, heh.

[DaringTremayne]

I don't think the problem has been solved....but I'm not sure

All I can tell you is that the only character sets I can support in-game are UTF-8 and LATIN1.

David

Very good - then we just need to convert from Latin2 (ISO-8859-2 in short) to UTF-8. I think on Linux a program called iconv can handle that conversion, I will check on that tomorrow (if it doesn't, others will anyway). Of course it could be simpler to ask the translaters to provide a UTF-8-encoded file, but perhaps that's just displacing the problem ???

In fact I could even write a Python script that converts from one encoding to another (going through Unicode), if that proves necessary.

BTW, anyone interested in the previously mentioned PDF file ? can I upload it somewhere on the Wesnoth server ?

[DaringTremayne]

I know nothig about solving this problem, he. Anyway I've wanted to submit Polish translation updated today but I believe we will have to wait 2 more days. Sorry for this but I have some issues to solve in my real life, heh.

Hi methinks,

Could you look over at the Polish file I have made in PDF and tell me whether that's how your messages should look like ? If so, how can I send it to you ?

Thanks,

Daring Tremayne