3 ms·

Whoa, that's interesting. Your description of the input scheme for the TI-82 sounds very similar to something I started building for a program editor meant to b

Whoa, that's interesting. Your description of the input scheme for the TI-82 sounds very similar to something I started building for a program editor meant to be used with motion sensors (https://www.youtube.com/watch?v=tztmgCcZaM4&feature=youtu.be&t=1m32s https://www.youtube.com/watch?v=tztmgCcZaM4&feature=youtu.be...). My plan was to always present a 3x3 grid with cells containing whichever language constructs were available from the current cursor position: after selecting one, the cursor position updates, and the options available in the cells update. I've tried to imagine using it many many times, but haven't been able to definitely conclude whether it would be usable or not...

derefr 10y ago
You're basically talking about an IME or software keyboard. For something similar to your idea, I'd suggest playing with the Japanese-language keyboard for iOS. For a bit of background: The two main Japanese writing systems are syllabic—each character represents a consonant+vowel pair, with ~15 possible consonants and 5 possible vowels, though not all possible combinations of those exist. So, instead of attempting to give you a ~70-key keyboard (wouldn't really fit on the screen), the standard Japanese IME for touchscreen devices instead displays one key per initial consonant (indicated with the character you'd get by completing the syllable with an "a"-vowel.) Tapping the key neutrally picks said "[consonant]+a" character; swiping in each compass direction instead picks one of the four other vowel options. And holding down one of the keys displays the options as a little plus-shaped menu.
- westoncb 10y ago
  Interesting. Yes, that is very similar to what I had in mind. The ideal solution for typing in Japanese is slightly different though since the things you're specifying are always comprised of two parts, allowing the system to always be two-step. I think I would have to go with a sequences of taps, since the things being specified could have many parts. Let's say you select "method definition", for instance. After that (assuming we're writing Java), you'd have to select an 'access modifier', 'return type', 'method name', and 0 - N 'method parameters', and so on.

derefr 10y ago

You're basically talking about an IME or software keyboard. For something similar to your idea, I'd suggest playing with the Japanese-language keyboard for iOS. For a bit of background: The two main Japanese writing systems are syllabic—each character represents a consonant+vowel pair, with ~15 possible consonants and 5 possible vowels, though not all possible combinations of those exist. So, instead of attempting to give you a ~70-key keyboard (wouldn't really fit on the screen), the standard Japanese IME for touchscreen devices instead displays one key per initial consonant (indicated with the character you'd get by completing the syllable with an "a"-vowel.) Tapping the key neutrally picks said "[consonant]+a" character; swiping in each compass direction instead picks one of the four other vowel options. And holding down one of the keys displays the options as a little plus-shaped menu.

westoncb 10y ago
Interesting. Yes, that is very similar to what I had in mind. The ideal solution for typing in Japanese is slightly different though since the things you're specifying are always comprised of two parts, allowing the system to always be two-step. I think I would have to go with a sequences of taps, since the things being specified could have many parts. Let's say you select "method definition", for instance. After that (assuming we're writing Java), you'd have to select an 'access modifier', 'return type', 'method name', and 0 - N 'method parameters', and so on.