4 ms·

This is good! I'm currently working on something related and I could really use this. May I ask how are you computing the shapes of rectangles? Is there some

• hcs 9y ago
  Thanks! There's no real optimization, everything is just done from the bottom up: a box is large enough to contain its rows, rows are large enough to contain their "cells" (which are just the boxes the next level down). There is some padding that is fixed, and some (the "handles" on either side of a row) that scales down with depth, though at a different rate than the text which is the main control over box size. Handles disappear entirely first, as it isn't possible to click on a row directly once it gets too small (instead a click will zoom in on the node). While pinch zooming some effort is made to try to keep focus on the same object, but that's just a matter of keeping the object in the center of the zoom in roughly the same place on screen. Happy to answer more questions here or by email (agashlin gmail), and I'd love to hear about your project.