KDE has published on his X account (how strange that sounds...) an article that bears the signature of David Edmundson. In his blog, the project collaborator has told us about new ideas for input methods in Wayland, and I think there is at least one for each type of user. The best-known input method is the virtual keyboard on the screen, although known is not always synonymous with used. The other is the Asian keyboard, and it is this that they intend to take advantage of.
The K project does not stitch without thread, so it is more than likely that we will see some or all of what is explained in Edmundson's article in the medium-long term future. From the looks of it, the idea of introducing new input methods in KDE it had been in the heads of its developers for a long time, but they will focus more on it at the moment in which Wayland is used by default.
Input Methods Discussed in the KDE Akademy
Edmundson took advantage of the Academy to talk about this topic. Although he says he doesn't use Chinese Japanese Korean (CJK) or Asian keyboard input, there are things he does use that could be beneficial from the InputMethod technology. Now they have a standard and can use it for more creative ideas like (visit Original article to see animations):
- Clipboard Connections. This would allow access to the clipboard from a text editor after using a keyboard shortcut. In the example, a text is pasted and a drop-down appears from which we can choose the desired option.
- Easier emoji input. Currently, to add an emoji to a text we have to press META + PERIOD, navigate through the emoji widget, click on one -copy it- and paste it where we need it. What we see in the example would allow us to put a colon (:) followed by a word so that it could search for them and be able to add them by selecting them.
- Input to add accents. I don't think I'm going to use it much with the Spanish keyboard, but they also envision an input method that displays single-letter options with multiple accents, something like what we see on mobile keyboards.
- specific translations. Using a text field, we could write in our language and what would appear in the document would be a translation.
- Voice to text. The last input method would allow you to dictate text. Edmundson says that's what excites him the most, and no wonder. Although there is speech recognition software for Linux, having it integrated into the operating system would make things much easier.
Actual state
All previous are examples, but it was talked about at the Akademy. Right now it is in such an early phase that it does not go beyond "playground", which is the word that developers use to define a testing ground in which to play without breaking anything. The repository is in this link. Thanks to that playground several problems have already been discovered.
So this is something we might see in the future, and I think we'll see it in KDE at some point. Without a doubt, everything exposed, except for the accent for obvious reasons, are things that I would like to have to a greater or lesser extent. Maybe I didn't trust the translator to write texts, but right now I only trust DeepL for this type of work. Yes it could be used for simple translations.
What of the voice to text It would go very well, like the emojis and the clipboard. In the latter, many times I copy something, it comes first and what I want to paste is something else, so I have to go to the system tray, find what I need, click on it and then paste it. The input method that they have given us in the example would make things much easier.
I only hope that KDE does not stitch without a thread is true, although it is certain that we will have to wait a long time to start seeing the results in a stable version. Patience.