Game Language

I know, that the game will be principally in English, but we have players from all the world.

I speak fluently Spanish and Lithuanian, so I can help to translate the game in those languages!

Edit 2 from my last message:

I think that I can be a good translator. I am the CEO & Founder of http://www.starbound.es ( The Spanish Starbound Community ) and I am translating everything, also doing the wiki. I'm making videos, having a big Spanish channel, I just want the game to be available in another language, there are children in other countries who do not speak or understand English very well. That's it.

 

Similar to this thread.
It seems translations are planned after release.

 

Translations will end up be possible indeed. I just don't use them, games on my fluent language are usually RUINED, they translate some stuff to funny things and can even create confusions to understand lore due translations done with "specific words" that are more simple on English.

 

I think they're planning to crowd source translations post-release.

 

All your base are belong to us

 

Yep, it's usually just better to play the game as intented in it's original language as translations might make a mess of things. But there will be multilanguage support down the road.

 

Lost in translation.

 

Why didn't you say English or English ? What're you biased?

 

Because he's clearly writing in English.

 

I am error?

 

No offense mate, but there's a pretty big difference between speaking languages and translating. In order to properly translate something, you'll have to be pretty much perfect at both languages.

 

And be creative for adaptation without falling within the cheesy territory. And know the context. And the eventual cultural differences plus their equivalents.

 

Also, I think anyone volunteering to translate is forgetting just how many damn words there are in the game - it's not like translating Minecraft which only has a few hundred things to translate. @Razey writes a lot.

An example: there were tens of thousands of words in game before we started writing NPC dialogue. All of that was just descriptions of items (seven for each one).

 

You are brazilian aren't you? Yeah, portuguese translations are bad...really bad.

Well, I actually see that as a reason for more people to be interested in helping to translate other than a reason not to. Sure, if the devs can arrange a group of brilliant translators who know about all the lore AND are fluent in both languages, then it won't be necessary. However, with thousands and thousands of lines of text I would be accepting anyone who is keen on helping as long as they can do better than google translate.

 

@Renlou i'm Portuguese, the original Portuguese translations are pretty Ok, but sadly games translating into Portuguese-Brazil instead of the original one just drives me crazy, Portuguese is original from Portugal but games translate with the Brazilian words, etc... that doesn't exist in here.

 

Few would argue that translation's aren't a good thing in general. They make the game available to a wider audience, increasing player popularity, as well as overall revenue for the developers, meaning in turn more content for the rest of us. They make players of all regions of the world, and all cultures, appreciate Starbound. Moreover, translated games are popular for YouTube, Twitch, and other streams -- while the streamer may know English well, a lot of their audience may not, and since everyone has to go at the streamer's pace, they won't necessarily have the luxury of slowing down to understand a foreign language passage. Being an indie game, word-of-mouth advertising (of which online streamer celebrities are a big part) is crucial for the game's commercial success, making a translated version even more valuable for players and developers alike.

However, before we rush for the "let's translate all the things!" hype, we should take a step back and have a rational look of what it actually takes to translate the game, and therefore, when is the right time to do this.

A. Translating the Game's Text:

The most obvious part, of course, is actually getting the game's text translated from English to another language. Traditionally, you'd hire an external company to provide all the translations, and send back a complete "translated package" when they are ready. This works well for AAA games (especially because there is not only text to translate, but also voice acting, graphical art, movie sequences, etc.) Unfortunately, this approach can take a long time and cost a lot of money, given the huge amount of text to translate for Starbound (e.g. I heard that a single race would have hundreds of speech and other text pieces for various contexts), making it potentially not viable for an Indie studio.

Alternatively, nowadays there are various online software vendors that allow crowd-sourced translations (essentially it lets multiple people sign up and translate various bits of text after comparing them to the original. The software sends the same bit of text to multiple people and compared the result, either automatically (for very short pieces of text, there is a good chance of several people providing an identical translation) or semi-automatically (present the top differences to a trusted reviewer). Then, the most popular and/or manually selected top picks are aggregated into a final translation table, and sent back to the developer for integration into the game. This approach allows to speedup translation of large amounts of text beyond what's possible for a single or few trusted reviewers, yet still allow quality control and oversight in order to pick the highest-quality translations, and reduce chance of abuse (e.g. someone deciding to be "smart" and inserting profanity into a translated string).

B. Modifying the Game Engine to support Translated Text:

Besides translating the game's text, the game code should be designed to handle translations. There are several aspects to this:

1. Instead of hardcoding English strings all over the source code, hardcode symbols representing a unique translation. E.g. in a place where a Glitch guard would warn a visitor to sheathe their weapon, imagine that originally there would be two lines: "Please sheathe your weapon, sir.", and "This is your final warning to sheathe your weapon!". These would be replaced by special keys such as "glitch.guard.sheathe_weapon" and "glitch.guard.sheathe_weapon_final". Then, an external file would exist for each language, and when the game is run, it would look at the player's preferred language, and use the set of translations mapped for that language, substituting each matching key for the value mapped for that key. Fallbacks would also be established to allow iterative development (e.g. if a translated Portugese value is missing, it would check whether a Spanish match is available, and if it's missing as well, then use the default English as a last resort). An example of such a translation mapping file would look as following (example is simplified, a real translation file would have hundreds of such keys and values):
   • en:
     • glitch:
       • guard:
         • sheathe_weapon: "Please sheathe your weapon, sir."
         • sheathe_weapon_final: "This is your final warning to sheathe your weapon!"
   • fr:
     • glitch:
       • guard:
         • sheathe_weapon: "S'il vous plaît rengainer votre arme, monsieur."
         • sheathe_weapon_final: "C'est votre dernier avertissement pour gainer votre arme!"
2. The game's engine must be designed to properly handle non-ASCII characters (e.g. Unicode). If the game uses a custom font (as Starbound probably would), a glyph would have to be drawn for each character, and the code should know how to map Unicode characters to the needed glyphs. The underlying libraries that store, load, and parse the text, must also properly handle Unicode strings -- something that for a relatively low-level programming language such as C++ (which Starbound uses) may be a non-trivial thing. And then throw performance & optimization into the mix - the translation library should not be slowing down the game, even when trasnlating a large amount of text on-screen.
   
3. Translated values must fit well on the game's screen without overflowing or distorting the UI. This is more relevant for special pieces of text which occupy a fixed space on screen and cannot stretch (e.g. part of the HUD that says "Life", "Energy", etc.) If a French translation of a word is twice as long as the English one, the word may not fit on screen, causing problems. Other types of text, such as codex entries and quest descriptions, are designed to take an arbitrary input and scroll as needed, making these types of text a bit more tolerant to variable length based on language. But there obviously needs to be a lot of QA to make sure everything looks well.
4. Other international issues than language itself must be considered. Some countries would write "January 3, 2013" as "01/03/2013", while others would write "03/01/2013". In some countries "1,234.56" is the right way to write "a thousand, two hundred and thirty four, and fifty six hundredths", while for others the right way is "1.234,56". Should the game take all of this info from the player's OS, or ask them directly? Would there be issues in multiplayer when two players want to make a trade? Etc. This is part of a larger topic called Internationalization (a.k.a. "I18n"), of which translation is just one part out of many.

The above points are just some of the issues needed to implement a game's translation. As you can imagine, it's a lot of work for everyone involved, not only translators, but also reviewers, developers, QA testers, graphics artists, etc. It will obviously take development time to implement and release.

The right time to translate Starbound:

While it's an obviously great thing to have the game be translated to many different languages, based on the above considerations, such a translation effort is best done as a "post-beta" thing, and perhaps even a "post-release" thing. We should wait for the game to be feature-complete, for the UI to be stable, for players to start playing it, for bugs to be ironed out, and for Chucklefish staff to have a bit of free time on their hands. Then we can talk about translating it.

 

Yeah, there are a LOT of differences between Portuguese(Brazil) and Portuguese(Portugal), it's quite hard for us as well to understand you guys. If Starbound gets translated to portuguese, we might need the two versions.

 

Wat. I was just saying he colored the words Spanish and Lithuanian in the color of their flags, but not the English word. I wasn't really serious anyways.

 

English, eh!

 

I am currently on a postgraduate course for translators and would gladly assist in translating the game into Polish. It would be good practice actually.