what is the right visualisation to represent code ? 

we like graphs because the graph representation could ideally remove all names, and create a colored-ball network of computation that we can look at and make wide sweeping grand generalisations about the codebase (RIP crystalfab)

i had a problem with graphs

a graph presents ambiguity in that a directed edge could mean the head is being called by the tail, but this is the exact opposite of the direction the computation result travels..

so we would need one for the order of function calls, and another for data flow, and taken separately they are non-intuitive, and taken together they are confusing

what if we went gödel and replaces function names with numbers ?