4 ms·
No loss in functionality, but it adds tens of milliseconds of delay when typing. Also, when using Emacs I often press and release the ctrl key without completi
by tlb 5y ago
No loss in functionality, but it adds tens of milliseconds of delay when typing.
Also, when using Emacs I often press and release the ctrl key without completing a control sequence, like when I'm thinking about scrolling but change my mind. I'd be splattering Js all through my source files.
- thaumasiotes 5y ago> No loss in functionality, but it adds tens of milliseconds of delay when typing. Well, no, it would add zero milliseconds -- and zero nanoseconds -- of delay. The difference between key press and key release does not affect typing at all. You're not waiting to see what happens after you press T before you go on to pressing I (or whatever letter you're following it with).
- sph 5y agoI disagree. Input-on-key-release feels very bad if you're used to it being on key press like all computers do. It feels like the keys are sticky and unresponsive. It's hard to explain unless you try it.
- tlb 5y agoWhen I'm typing normally, keys are down for around 60 mS each. Having the software respond at the end of that 60 mS instead of the beginning adds about 60 mS of delay.
- rgoulter 5y agoI can't discern the difference between those tens of milliseconds of delay. Worth considering the time (and effort) saved from moving the hands to the modifier keys. It's nice being able to have the hands remain on the home row. EDIT: rather, there's hardly any speed lost (if any); but there's a big amount of comfort gained from using home row mods. > Also, when using Emacs I often press and release the ctrl key without completing a control sequence The behaviour is customisable, but e.g. if you're holding the tap-hold key for longer than a 'tap', then it'd still register as a 'hold' rather than sending a 'tap'.