Choosing the right language for web applications is not quite an easy task. There are many source of information and you have to figure a good algorithm for making the best decision. As I didn’t found yet the “right” implementation I decided to describe the algorithm.
If you want to minimize the number of bugs reported for your web application you should know that codepages are domain of the past. If you want to create multi-language web content all you have to remember is that you should use UTF-8 encoding everywhere
When I’m talking about keyboard shortcuts to things like Ctrl-S or even one key shortcuts like ‘A’ (used to Archive an email on Gmail). On the other side the mnemonics, known as hotkeys or the underlined letters in menus or other GUI controls are clearly something that should be localized and I not going to argue about this.
On Windows, keyboard events are reporting scan codes, virtual key codes and characters. If you want to deal with keyboard shortcuts and be able to use them in an international context you have to use the virtual key codes.
If you’ll ever try to use automation for testing globalization readiness on applications you are going to discover that the keyboard input API does not have any function for entering Unicode texts. If you’ll try to use VkKeyScan for converting an Unicode character to a virtual key - in order to map it to a keystroke - you’ll discover that the Windows Keyboard Input API will return you nothing because current keyboard layout is not able to produce the requested character.
One common error in internationalization is to think that all shortcuts must be translated. People forgot that one of the major advantages of the shortcuts is that they are memorized their position on the keyboard and not by the sign on the key.