Re: Build 159


Thank you Dave. The delay in adding the gesture to toggle is less about choosing what the best gesture is, and more about the fact that the full screen mode operates very differently compared to all other modes, and switching them on-the-fly is simply not yet supported in the backend code, so to speak. On-the-fly switching introduces a host of new complexities that I'm hoping to tackle soon. The full screen height uses the delayed textfield interaction mode, whereas all other heights use direct textfield interaction. One idea here is to make the full screen height also use direct textfield interaction, and have the contents immediately reflected on the textfield with all the benefits and drawbacks that would come with this such as the annoying "end of document" announcement when flicking down. And possibly have a setting for advanced users where they can toggle between direct and delayed textfield interaction modes independent of keyboard height. If these two things get decoupled then switching keyboard height on-the-fly would be trivial to implement. Hope that makes some sense.

Regarding tap to read back: This is really only meant for the full screen mode. In all other modes where you can interact with the app above, what happens if you tap inside the application area is completely out of our control as a keyboard. Typically, if you touch the text inside a textfield iOS will read the entire contents back I believe. The new FlickType gesture seems to kind of work in the minimal mode simply because there is a 1-pixel high visual announcement area that you will cross as you drag up, but you should have no reason to use that over the standard iOS way of reading back your text anyway. And regardless of mode, the read back from FlickType still suffers from the issue where iOS does not allow us to know the entire contents of a textfield, only the current sentence or so. In full screen mode, which also has the delayed textfield interaction, you can read back all the text you have written because FlickType keeps track of it all until it's dismissed and then inserts it into the textfield. Also regardless of what gesture you use to invoke the FlickType read back action, the text that will be read back is the same. I am sure my explanation in the change log about this new gesture was not great, and this current long explanation might make things sound much more complicated than they are from a user's perspective. To recap, the new 1-finger gesture is meant for the full screen mode, and you can simply imagine it as being able to sort of touch your text that always sits right above the QWERTY row. Hope that makes sense, and I'm sure someone else will be able to explain it better from a user's point of view 😎

- Kosta

Join to automatically receive all group messages.