Will Crichton

A hypothesis: for live programming to scale beyond toy programs, generic low-level visualizations aren't enough. Instead, we need Live Literate Programming: better tools for creating *program-specific* views, which communicate our higher-level understanding.

Bearish on program visualization right now. People have been trying to vis code for half a century, and nothing has worked. All we have are Doxygen inheritance diagrams. My guess: without a more principled understanding of program comprehension, we can't build effective vis.

I /think/ when you say "program vis" and "program comprehension", you refer to a globally-applicable, automatic process. Is that right? If so: That seems like a big assumption, when "program vis" could instead be a (fractally-automatable) crafting process like programming is now.

I am imagining a generic system, but generic != fully automated. e.g. data vis has a grammar of graphics that enables rapid construction of novel vis types. Today, all program vis just starts w/ standard web tools or canvas.https://twitter.com/wcrichton/status/1253758394982256640 …

Yeah, grammar of graphics for program state viz is kinda what I imagine we need

I think that's a very interesting prompt! Just wanna note a difference between the situations. For data vis, the data is given and you need to visualize it. For program vis, you can change programming itself. Prog&vis could be intertwined in a way not represented by GoG analogy.