What if text editors supported drawing a decision tree inside of code instead of writing a sequence of if-statements; or typing nicely formatted math equations; or embedding editable graph diagrams; or embedding image editors? In general, what if code editors could embed visual editors in text? I’ve been exploring this question through building Polytope, a code editor prototype that combines text and visual editors.

In a text editor, we navigate through code by moving a caret (Ꮖ) using our keyboard and mouse. In Polytope, navigation has to work between both text and visuals. Can we adapt caret navigation to visual editors, to create a bridge between navigating text and visuals? In particular, can we translate index-based caret navigation into spatially-based caret navigation for visuals?

This talk is a follow up to my PX/22 talk: “let chart = ⊥; let song = ♩; // Embedding Visual Languages in Code” (https://2022.programming-conference.org/details/px-2022-papers/7/let-chart-let-song-Embedding-Visual-Languages-in-Code).