I wonder if an approach like KKeyboard with larger but statically combined keys leads to faster typing than the current approach with smaller but dynamically "combined" keys.
[*] In reality, the context is modeled using a simple Hidden Markov Model with a much smaller effective context window that could not associate "Paris" and "France." But you get the idea.
[0] https://forums.macrumors.com/threads/impossible-to-type-okee...
* Does this still expect you to hit every key but some of them need multiple taps?
* Are they doing fancy autocorrect-like magic to decide which letter you meant, and if so why use this instead of taking it one more step and using http://minuum.com/ ?
* Or is it something else?
I guess that's better for precise typing, but for normal prose it's probably faster+easier to just type blindly and let the machine figure out what you mean.
I noticed "copyright info goes here (c) 2025" - which you might want to update!
It's still 2025, so that's fine :P post-Berne Convention, there are no forms required for copyright protection to vest, merely fixing the work upon a medium is sufficient.
teach•1h ago
Cool idea though.
mgr86•1h ago
https://archive.is/gW1rO
QWERTYmini•1h ago
really appreciate the help!
QWERTYmini•1h ago