Monday, October 24, 2011

The Specifics of Website Translation


Some may think that translating a website is no harder than translating a book or a manual: both include titles, texts, more or less the same footer information on every page, and a "table of contents" ("menu" for the website). However, if you get the chance to translate a website, you will soon realize that it is a tougher task than what some may expect.

Of course, at the end of the day, your job is still to translate words and sentences from one language to another, but unlike plain text translation work, you cannot just open up a web page and start translating everything that you see. 

If the webmaster has provided you with all the texts that you will need to translate in a plain text format, then you are a lucky person and you can stop reading this article.

For now, let me list some of the main differences between website translation and translation of "ordinary" plain text.

1. Books are made of paper, letters, and words. Websites are made of HTML

Web pages, at least the ones reaching your browser, are all made of HTML. Even if their names end up with ".php" or something different from ".HTML", the final content is always HTML, which is formatted text and not plain text. It means that if you\'re not translating web pages from code editing software (such as Notepad++ or Komodo), then you are doing things wrong. 

Rather, you will need to replace the text directly inside the code, using the previously mentioned software. As you can imagine, making changes inside the code can be quite tricky, as you may accidentally "break" the page. When you do it (I mean translating, not breaking the page), be very careful with tags and formatting. If you move or - even worse - remove some tags, the layout of the page may differ from the original version.

In the worst of cases, the whole page may even become unreadable or produce error messages. If you don\'t understand some characters or tags, just leave them as they are.

2. Web pages have a lot of "invisible" text

A web page is not limited to its visual content. Indeed, a lot of textual content is not visible when you just check the page on your browser. This is why, once again, you have to look at the code rather than what your browser displays.

You always have to check everything from the header to the footer of the page: Did you translate all of the meta tags, such as the page\'s description (the piece of text which will usually appear in search engines when people look for your website)? If mails are sent by the website, are these mails also translated? What about the alternative texts for images? Error messages?

Check the code again and again to make sure there's nothing you’ve forgotten.

3. Some words are not meant to be translated

Some parts of a web page don't need to or must not be translated. Before starting your translation, ALWAYS ask the webmaster if there is any content that is NOT to be translated. Some parts of the page (usually non-visible content) must be exactly the same in all languages to ensure that the website will work correctly. For example, tag attributes such as name="xx", id="xx" or even sometimes value="xx" should not be translated. If you didn't understand a word of the previous sentence, then you should be extremely careful and, if possible, learn a few elements of HTML, at least to understand how web pages are structured.

4. The problem of image localization

Web pages are made of formatted text, colors and images. Images including text are a concern for anybody who has been translating/localizing software or websites.

The first (important) thing to do is to get the image template file used to create the original image. Once you have it, you can edit the text layer with an image tool. Usually, Adobe Photoshop or The Gimp will do it, but if you don\'t know which software you should use, ask your client. 

Once the new text is implemented, there are two possibilities:

-The new text fits the image: You\'re done!

-The new text is too large: You will have to call your client to know what to do. Should you use a shorter but less accurate word? Reduce the font size? Increase the size of the button, which could bring about the necessity (for the webmaster) of rewriting some parts of the code? In any case, do nothing without the approval of your client; explain the problem as clearly as possible and ask the webmaster\'s advice.

5. What looks fine in your editor may look awful in your browser

After you finish your translation, checking the new page in a web browser is always a good idea. If the layout looks different, it certainly means that you’ve moved or deleted some tags by mistake. Always keep a version of the original page so that you can compare the code and find your mistake.

But don't forget: having a correctly displayed page doesn't necessarily mean that everything was translated properly.

There are a few other points I could mention here, but they are usually specific to certain categories of websites, and I believe the client would give you enough documentation in such cases.

As you can see, website translation differs significantly from "ordinary" translation. It can be quite tricky and you will always have to check everything extra carefully.

If you know nothing about the structure of an HTML page, then I strongly recommend that you read some introductory articles. You don't need to become a webmaster yourself, but you should at least understand the structure of an HTML page, and if possible understand the meaning of some "Webmaster slang" words as well, such as "meta tags", "tag attributes", and "page footer".