Re:Dijkstra's Algorithm
by Silhouette on Thursday November 10, 2016 @04:12AM
(#12918872)
Well, I'm sorry if Dijkstra would have been offended, but vast amounts of human experience with teaching and presentation methods say pictures work, animations work, and interaction works. Awful presentation is unfortunately the accepted norm within academia, at least in scientific and mathematical disciplines, but that doesn't mean everyone else has to sink to the same depths when trying to pass on useful information or convey understanding.
https://www.cs.utexas.edu/~EWD/ewd12xx/EWD1255.PDF
At least one instance of using pictures to illustrate things.
Your quote is about being opposed to weak (informal, hard to show correctness or make consistent across the range of things he wanted to prove) proofs by diagramming, where he wanted more formal reasoning.
The man was no fool, in at least some of his presentations I've seen (videos) he also used diagrams and pictures to illustrate things. They're necessary on occasion.
EDIT: Typos from trying to comment from my phone.
Dijkstra must be turning in his grave. He was strongly opposed to the use of pictures in his papers.
I can't find a direct quote, but https://www.cs.utexas.edu/users/EWD/transcriptions/EWD10xx/E... says:
"But this historical explanation is no excuse for the fact that today Euclidean geometry, with all its known defects, is still taught as the prototype of a strictly deductive system. It is not. Its reliance on pictures is quoted as evidence for the value of “geometric intuition”, while in fact it is only a symptom of the defects of Euclid's axiomatization. (In passing I would like to add that I see more and more evidence that it is precisely the informality of the pictorial argument that defies the development of heuristics by means of which such arguments can be designed in an orderly fashion. This is in sharp contrast with the highly effective heuristics that could be developed for strictly formal environments.)"