Text In Rectangles: The UI Desideratum

I think about user interfaces frequently, usually for a few hundred to a few thousand milliseconds at a time. And my thought is: This is very pretty, and it’s not what I want.

A computer monitor is a rectangle, and everything a computer does maps to ones and zeroes which can be most compactly represented to humans by strings of text. The root of all user interfaces is displaying text in some rectangular shape. And this has been a solved problem since the 70s. (It was a small rectangle, at first.)

There are things computers do that don’t involve text in rectangles. Movies, music, video games other than Dwarf Fortress and Nethack, etc. But these all evolved from non-computer things; they’re efficient to implement with a traditional general-purpose computer, but they’re best thought of as a virtual environment; computers simulating TVs, radios, or game boards, not computers being computers.

Every frustration I have with modern interfaces stems from this problem. I want to interact with a stream of text. One character or one line at a time is inefficient, but humans have been looking at pages of text for thousands of years. We’ve gotten pretty used to pages of text as a medium for compactly representing information. Anything that gets in the way of that has a high burden of proof.

I’m not against graphics or even fancy visual effects on principle. A painting, or even a chart, is obviously not something that should be represented as a stream of ASCII characters. What I am…