In parts one, two, and three, I tried to explain 1.) just what is open source software, 2.) what is open content and 3.) why both are in search of better translation tools. Now it’s time to look at how open source software programmers and open content creators think about translation.

First off, I think it is useful if we divide ‘translation’ into three different categories.

1.) The translation (localization) of software.

This refers to making a piece of software (like Firefox, for example) available in other languages. To do this, the developers of the software program divide all of the text found within the program into “strings”. Each ’string’ is a short collection of consecutive words. An example of a string in Firefox would be ‘Save Page As …’ which is found under the File menu pulldown. In Spanish, that string is translated as ”.

2.) Translation (localization) of website user interfaces.

Blogs and most websites these days have two types of content. There is the text that always stays the same (’interface’) and the text that is frequently updated (’content’). For example, this post that I’m writing now is content while the menu items above that say ‘Begin’, ‘About’, ‘Photography’, etcetera, are ‘interface‘ because they always stay the same; they are part of the website’s design. Translating these two different parts of a website (interface and content) involve two different processes. To translate the interface of a website, we use the same process as translating a piece of software like Firefox. We divide all of the text of the website’s design as ’strings’. We then need to compile a list of all the strings that need to be translated in order to make the interface available in another language. The most common type of file which lists a series of strings and their translations is called a PO file. (The other standard way of doing this is with XLIFF files.)

One of the strings from WordPress is “Last 15 Posts”. That string can be found in the WordPress POT file and can be translated into any other language. A Bengali translator, for example, would take that POT file and add the translation of “Last 15 Posts” as “সর্বশেষ ১৫টি পোস্ট”.

Here is a list of various programs that can be used to create and edit PO files.

3.) Translation of content.

Notice that it doesn’t just say ‘web content’ because any type of content that can exist online can also exist offline. For example, one of the conference attendees, Tomas Krag, is the author of a book on “Wireless Networking in the Developing World” which is available both online and in printed book-form. The book has already been translated into Spanish and French, but Tomas would obviously like it to be made available in as many languages as possible.

When we talk about translatable content, we’re really talking about just two types of things: text and voice. Images and video are not translatable unless they involve text and voice (which video usually does.) When it comes to multi-lingual open content websites, unfortunately, there aren’t many success stories. Translating content like a blog post or a five minute YouTube video is often more time-consuming than writing the original post or making the original video. And while being the author of good web content can turn you into a web celebrity, translators of that content are rarely given any recognition.

There are three open content, multilingual websites that stand out as success stories. First is Cafebabel.com, which dubs itself as “a European current affairs magazine: comment and analysis from Europe in 7 languages.” The website - and just about all of its content - is available in English, French, Italian, Spanish, German, Catalan, and Polish.

The biggest open content success story, without any doubt, is Wikipedia. The open content encyclopedia is available in more than 100 languages. Some languages, such as Aymara, have less than 1,000 entries while others like Galician have over 10,000 entries. One of the really great features of the website is that when you go to any single entry - for example, Karl Marx - on the left-hand side, you see a list of all the other languages which the entry is available in on.

This is the same basic feature we use on Global Voices to let someone know on the main site that a particular blog post is available in other languages. For example, when I recently published my interview with Cristina Quisbert of the Voces Bolivianas project, Global Voices’ kind volunteers translated it into French, Portuguese, German, Bengali, Spanish, and Malagasy. Note that not only was the post translated into all those languages, but the video was also subtitled in Bengali, English, Spanish, and Portuguese.

Each individual translator from each of these three open content projects (Cafebabel.com, Wikipedia, and Global Voices) has their own methodology and their own toolset to help them translate content from one language to another. Global Voices translation coordinator (Lingua Director), Alice Backer, recently sent out an email to the mailing list asking for sample workflows. The responses from some of the translators reveal just how varied the translation process is depending on which languages you’re involved with. In the next post in this series I’ll take a look at some of the tools our Global Voices’ translators are currently using and some of the obstacles they’ve encountered.