Windows XP Internationalization (i18n) Tips

Windows ^XP continues the good work Microsoft started with Windows 2000 (see International Features of Windows 2000).

For anyone working in a multilingual environment, Windows XP is a superb platform. Here are some of the things you can do:

• Maintain a user interface in your native language for all language tasks
• Handle text in any language, using built-in tools like Notepad
• Convert from native text to Unicode and back using Notepad
• Emulate localized versions of Windows
• Have full access to Asian input
• It has the same range of multilingual features that make it very useful for anyone involved with more than one language.
  This document is shows how to access some of these useful features in Windows XP.

Before describing Windows XP features, let's review what is available in Windows NT4 and Windows 2000.

Windows NT4 has a Unicode code-base, which means that it can support more than one language.  With Asian support, for instance, you can display Chinese, Japanese, and Korean data on English Window NT4.0. With the addition of the appropriate Input Method Editor (IME), you can also enter data in these languages.

With the appropriate fonts, you can display and edit almost any language except bi-directional language (Arabic, Hebrew) in Notepad.  Notepad also offers the option to save data as Unicode (UCS2).
Beginning with Windows 2000 you can also display Arabic, Hebrew, and some Indic languages in notepad.

Windows 2000 offers a wide variety of language features that will be helpful to the internationalization industry. Some of these features are described in International Features of Windows 2000:

1. Single executable
2. Full Unicode support
3. Ability to select language per-thread
4. Multilingual documents can contain mixtures of Asian, European, and Middle-East languages
5. Selectable user interface language
6. Ability to change user interface language without rebooting
7. Multilingual and Unicode enhancements to Notepad
8. Other features

Setting up Windows XP for International

Installation

Windows XP handles support for other languages differently from Windows 2000. During installation of Windows 2000, there is a dialog offering support for several languages and character sets, such as Cyrillic or Eastern Europe.

Windows XP automatically loads support for all the "easy" languages that are offered in the Windows 2000 installation, and there is only one check box to choose during install. That check box is for Asian languages, which are not loaded by default because the font files are so large. To add support for Japanese, Korean, Simplified Chinese and Traditional Chinese, check this option. If you decide to do it later, refer to the following section.

Select the control panel (it needs to be enabled in order to show up in the start menu):

Click on the Regional Settings icon:

Selecting the Regional Settings icon will bring up the following dialog box:

The general tab handles locale settings that are not specifically related to language. So let's move on . . .

Click on the second tab, "Languages", and the following dialog appears:

Note that Windows XP has only two check boxes for languages; while for Windows 2000 there are dozens.

• Install files for complex script and right-to-left languages (including Thai)
• Install files for East Asian languages

The reason there are only two check boxes is that all the "easy" languages like Western European, Eastern European, Cyrillic, and Greek have been installed automatically. These are all single-byte languages, but not all single-byte languages are "easy". See Complex Script below.

These include Japanese, Korean, Simplified Chinese and Traditional Chinese. These are also called "double-byte languages", because their non-Unicode representation takes two bytes. If these were not selected during Windows XP installation, you can do it in this dialog.

If you check the complex script box, you will get Arabic, Hebrew, Thai, and some Indian languages. They are called "complex" because the layout of the text is indeed complex. For Arabic, for instance, the shape of a character depends on the characters (or lack of them) on either side of the letter. Thai also has a complex method for drawing the characters.

We can add input locales, such as Cyrillic keyboards and Asian input method editors (IME).
Select Details in the section labeled "Text services and input languages". The following dialog appears:

Select the appropriate input language, and method.

Options for switching between input methods (languages) are displayed by selecting "Language Bar".

Other keyboard options are available in the "Key Settings" dialog:

The "Languages" drop-down has two main uses:

• Change the keycaps for another language, such as Russian
  
• To add an Asian Input Method Editor (IME) which allows Chinese (Han) characters to be generated on a normal keyboard. For example, let's add a Japanese IME:
  

The resulting Regional Settings Language tab looks like this:

There are a multitude of settings for Japanese alone, which would be a good topic for another web site. Here is one of the option dialogs for the Microsoft Natural Input 2002:

Note: For non-Japanese speakers testing applications with Japanese text, the IME is not absolutely necessary. On the other hand, it is easy to learn, and a good way to enter Japanese text using a standard English keyboard.

See also International Windows 2000 IME

Microsoft has changed the description of the first option on the advanced tab, due to widespread confusion. The author can testify that may people were confused by the "Set Default" tab on the Windows 2000 Regional Options General tab. After explaining it countless times to users around the world, it is gratifying to see a better description on the Windows XP dialog:

"Language for non-Unicode applications" is the key phrase, and the dialog goes on to say:

This is a very important concept, and it will save a lot of grief when you want to look at legacy applications (or brain-damaged applications) that do not handle Unicode.

In general, it is not possible to know which programs support Unicode and which do not. Therefore, if you are working in Korean:

• some of your programs will show Korean just fine (e.g. Microsoft Office, Notepad).
• other programs will show either question marks or garbage, whenever a Korean character is displayed.

The solution is to set the "System default code page" to Korean, which is the Windows 2000 terminology that describes this dialog box. In Windows XP you set the "Language for non-Unicode programs" to Korean, as shown above, and then Korean will show up correctly in all applications (assuming you have installed Asian support).

This setting is required for Arabic, Hebrew, Asian languages, Russian, and any language whose alphabet is different from Western European.

For Windows XP, there is a separate CD or DVD with files labeled Microsoft^® Windows^® XP Multilingual User Interface Pack, Discs 1-5.

Languages, code pages, and the corresponding folder names are as follows:

• Here is the installation dialog for disk 1:
  

For further discussion of Multilanguage Windows, see International Windows 2000.

According to Microsoft, a Windows 2000 Multilanguage installation contains about 94% of a true localized operating system.

In other words, an English base installation, with Japanese multilanguage files installed, acts like Japanese Windows 2000 within to a 94% level.

Windows XP brings that number up to about 98% (these numbers may have changed since we heard about them).

Windows 2000 is a great platform for international testing, but Windows^XP is even better.

On the basis of its international features alone it is a true winner