A couple of comments here mention using this in VR. Fwiw, years back I played a bit with shallow-3D UIs for software dev. Shallow like within a few cm of a laptop display, to minimize VAC eye strain for all-day use. Think more being able to layer and draw in color, but in 3D, rather than waving arms in a room. The 3D can be wiggle 3D, or perspective from webcam head/eye tracking, or stereo from shutter glasses, or XR…
I have a working fully 3D glyph based text rendering system I can't seem to get people to look at. It's this. Every character is a 3d placed quad, instanced rendered, so you get tens of millions and then some. They are individually addressable and mutable like any polygon. I use it to render entire GitHub repos in one go. I have two versions, native Apple and web. Web has the basics of an ide setup. Would love insigh…
Nifty.
Some scattered thoughts...
First-time visitors easily bounce away if there's friction. My experience was: After reloading the page to reread the "load a repo" message, it took a search to find it. Two-fingers unexpectedly panned not zoomed. I consulted the keyboard shortcuts found earlier. My scroll panned. Keys zoomed, but speed was limited by key autorepeat. No response when clicking on shortcuts, suggesting they can't be changed. I eventually noticed shift-two-finger zoomed.
So perhaps add an initial popup, as with browser games, giving basic orientation/keys? And perhaps have some default repo already loaded?
Is antialiasing turned off? The text is less readable than I'd expect, and there's flicker during zoom.
I didn't find the usual github-icon link to the repo. Add it?
Perhaps try google/bing maps-like zooming towards cursor location, rather than always towards center?
Perhaps clamp zoom-out and scroll so one can't reach no-text-visible blank black screen? Perhaps clamp zoom-in earlier, so "slam inward" stops short of the current one-line-high screen?
Clicking README.md brings up a 3 column render, which is too small to read on my screen without further zooming. I'm unclear on what I'd like instead.
A big picture observation. Early phone apps did skeuomorphic UIs - a calendar app might render a wire binder spine, and animate paper page turning, with sound effects. It helped onboard a user population unfamiliar with phones. We don't do that anymore. It'd be silly. Most VR UIs seem to me skeuomorphic transients. Why hobble ourselves by unnecessarily importing the severe constraints of a Euclidean physical world?
So for example, file presentation might combine content-aware and location based scaling. A README might show up with the header lines enlarged, as in some md editors, and perhaps with fisheye-like distortion (when at the top, the top is rendered normally, but with increasing shrink below). Or mix in outliner-like expand/collapse. And so on. Whatever dynamic geometries are task helpful for the human. Once text colorization is working, it might be fun to work little exercises (make videos?) - with files X and task Y, how might they be most helpfully presented? When skimming agentic markdown as it scrolls by, I keep wishing for a couple of cloud-fast small models that watch it with me, and adjust the text size/color/annotation/etc to guide my attention to notable bits.
In the vein of making functionality easy to explore, perhaps give diff a default pr?
It's neat to see a repo slurped into something like this. Thanks for sharing.