# Graphical flow for programs

Can we model more general program flow in a similar fashion to Monocl?

# Limitations

The idea of graphical programming languages is linked with theDeutsch limit (named for noted programmer L Peter Deutsch, not physicist David Deutsch FRS, though perhaps he could come into play later):

The problem with visual programming is that you can’t have more than 50 visual primitives on the screen at the same time.

# Automatically create visual interfaces

Here's an idea: assuming we have enough text mining pixie dust (on corpora of linux man pages, and stack overflow questions/forum posts about linux commands), it might be possible to do:

user:~\$ make-gui-for ls --output ls.py

# Feature: Build infra for generating and displaying graphs.

E.g., we can generate graphs based on code flow.

(defun triangle (n)
(if (equal n 0) 0
(+ n (triangle (- n 1)))))


This would then be related to the visual code walk through feature described below.

# Feature: Visual code walk through

Ray is working on a visual code walk through. This should be seen as another interface to the same basic underlying information, sort of like how Org Roam is the main interface to the data served by Org Roam Server.

## General evaluation strategy for these demos:

• ‘Would anyone want to use this?’

• E.g., in the case of Emacs "learn X in Y" demo.

• If there is interest, work up to covering the HtDP book

• MAUDE framework. :: You describe your programming language using rewrite rules in K. They define tools to auto-derive rules in K.

• Program slicing :: ‘Galois connection on the traces’. This allows you to find where bugs appeared. People tend to look in the most recent. Imagine a call-graph of all the variables, so it gives you a minimum trace, showing where your bug can be found.