Although both macOS and Windows support multiple languages, if you are a multilingual user, especially one who frequently switches between windows and documents, and each environment requires the use of more than two input languages, the nightmare of switching input methods begins.
First, on a pessimistic note, this problem may never be fully resolved because the need to switch between multiple input methods is relatively low-frequency. After all, the number of people who frequently use three or more input methods is quite small.
Even people who use three languages might not use three input methods. For instance, when typing in German, most of it can be done with an English keyboard. Characters like ä, ö, ü, and ß can be substituted with ae, oe, ue, and ss.
Although switching input methods is a hassle when typing in multiple languages, two other issues truly frustrate users:
- The first issue is realizing after some time that the wrong input method is being used, requiring the user to delete previously typed characters one by one and start over. This often happens to users who cannot touch type (or are typing from a draft) because they focus on the keyboard (or draft) and neglect the screen; or to fast typists who produce a string of characters before they realize their mistake.
- The second issue, following from the first, is that users often cannot quickly predict which input method will be next when switching. Sometimes they accidentally skip over it and have to cycle through again.
These two issues severely impact the continuity of typing. If they occur repeatedly, the user's motivation for writing can be entirely depleted.
To solve this problem, we need to understand the working principle of input methods. Besides direct input methods using Latin letters, most other input methods have a character (or word) library, such as Chinese and Japanese. When an input method is selected, the binary file of the character library is loaded into memory, waiting to be called. The content we input with the keyboard is actually a set of filtering rules, and the outputted content is the result filtered by these rules from the character library. Understanding this principle, a solution becomes clear:
To address the second issue—users cannot quickly predict the next input method—systems can allow users to customize a set of shortcuts for each input method, such as Control + H for Chinese, and Control + J for Japanese. Alternatively, Control + Space can switch between two commonly used input methods, such as English and Chinese, while a custom shortcut can be used for less frequently used methods like Japanese. This way, users will have an accurate expectation of the input method switch result without any memory burden.
If the second issue is resolved, the first issue becomes easier to handle. Since the principle of input methods is a set of filtering rules, the system can allow users to input the filtering rules first. If the input method is wrong, they don't need to delete the previous content but can directly switch to the target input method. The previously entered content remains valid under the new input method. Wouldn't this solve the continuity issue in typing?