Trent W. Buck <trentbuck@gmail.com> wrote:
Last time I looked the iPad's onscreen keyboard didn't even do dvorak...
But there are overwhelmingly more Chinese language users than dborak users, hence an enormous economic incentive for phone and tablet vendors to support the former.
[Concerning computer keyboards, not handsets...]
AIUI hanzí input methods (inc. pinyin) basically have you spell it out more-or-less phonetically on qwerty, then allow you to select amongst homophones.
Is there sufficient detail in the Romanized transcription to give you the tone as well as the speech sounds? Just my curiosity taking over...
Hangul and (I think) kana have few enough graphemes that you basically have one key per grapheme, plus a few dead keys (a la diacritics in western europe).
Incidentally, Hangul keyboard layout has *always* been vowels on one side, consonants on the other. Take that, qwerty!
It's a good idea. Qwerty was, after all, designed to slow down typists to prevent mechanical difficulties in early typewriters. It has also been hypothesized as not accidental that "typewriter" involves only keys along the row above the home row, i.e., it's easy to type quickly for demonstration purposes.
Finding a phone with Chinese printed on it and with original documentation might be a little more difficult, but if the user is prepared to work with a device that has English labels printed on the controls then I would expect it to be entirely a matter of software.
I don't think that's an issue for Mandarin; you just use qwerty. For kana and hangul, of course, have alternate graphemes on the keys. If you can touch-type, it's not TOO hard to simply remember the mappings without looking at the keys -- at least if you can see the graphemes appear on screen. I struggle a bit to type Korean words accurately when I only get diamonds on the screen ;-)
That's entirely understandable. Typing Korean on machines without the right fonts installed is impressive, but I wouldn't recommend it. for security, one would have to be able to type passwords/passphrases without seeing the graphemes, however. For my part, I can select a Chinese table in my braille display software, type LANG=zh_CN date at the shell prompt, and get output that may or may not be intelligible to someone who reads Chinese braille. In fact, my recollection is that at the console, languages requiring large fonts aren't supported anyway, due to the kernel's limited font table. There aren't any Japanese or Korean tables at the moment.