Somewhat off-topic, it'd be cool if an operating system could offer different "input methods" for input fields (not just on websites). On phones one can have different graphical keyboards, e.g. for emojis, or to insert GIFs, or hand-gesture. On the desktop it's just the physical keyboard, but imagine being able to add a layer between the keyboard and textarea. You're on github and want to quickly edit code? Tell the OS to launch IntelliJ and store the output of IntelliJ in the textarea (although, I guess expecting IntelliJ to offer advanced code completion without checking out the git project would be too much of a headache...). Don't like WordPress' editor? Launch Word and have its output entered directly as a WordPress "article" (although that involves a step where "Magic happens", which this article shows would need a bit of coding...)<p>Of course the above might be unnecessary complication where copy-paste works fine. Already rich text editors on e.g. Facebook's website are magical, one can just copy image data and paste it and it would attach the image to your FB post/message, geez, we didn't have that in the 90s!