4 ms·

To me, the strange thing about this concept is the emphasis on textual code. There is still a screen and a keyboard, they are just projected or on paper. I thin

by amatic 7y ago

To me, the strange thing about this concept is the emphasis on textual code. There is still a screen and a keyboard, they are just projected or on paper. I think I would enjoy more something like a cable-patch interface. I would draw an integrator on a piece of paper, draw a line to a plot window, also just a rectangle on that paper, draw and an incoming signal from a generator to the integrator, then play with that. Something more like Bret Victors previous demos on dynamic exploration and simulation.

Animats 7y ago
I think I would enjoy more something like a cable-patch interface. I would draw an integrator on a piece of paper, draw a line to a plot window, also just a rectangle on that paper, draw and an incoming signal from a generator to the integrator, then play with that. That's called LabView. Widely used with instrumentation, or for simulating instrumentation. The trouble with programming by wiring stuff together is that it doesn't scale. It's been tried many times. The Blender game engine worked that way, and you'd end up with a maze the size of a tabletop for a modest game. At some point you need abstractions, and that model doesn't lead to them.
- amatic 7y ago
  I was going trough wikipedia after your comment, browsing different visual programming languages and paradigms. I suppose you're right that they don't scale in the same way as textual languages. Maybe someone finds a good set of constraints to make them scaleable and the elements abstractable. Anyway, they do have niche success (if not in general programming); and some of Bret's explorable explanations ideas [1,2,3] seem to point away from textual programming. [1] http://worrydream.com/#!/ExplorableExplanations http://worrydream.com/#!/ExplorableExplanations [2] http://worrydream.com/#!/InteractiveExplorationOfADynamicalSystem http://worrydream.com/#!/InteractiveExplorationOfADynamicalS... [3] http://worrydream.com/#!/LadderOfAbstraction http://worrydream.com/#!/LadderOfAbstraction

by amatic 7y ago

Animats 7y ago

I think I would enjoy more something like a cable-patch interface. I would draw an integrator on a piece of paper, draw a line to a plot window, also just a rectangle on that paper, draw and an incoming signal from a generator to the integrator, then play with that. That's called LabView. Widely used with instrumentation, or for simulating instrumentation. The trouble with programming by wiring stuff together is that it doesn't scale. It's been tried many times. The Blender game engine worked that way, and you'd end up with a maze the size of a tabletop for a modest game. At some point you need abstractions, and that model doesn't lead to them.

amatic 7y ago
I was going trough wikipedia after your comment, browsing different visual programming languages and paradigms. I suppose you're right that they don't scale in the same way as textual languages. Maybe someone finds a good set of constraints to make them scaleable and the elements abstractable. Anyway, they do have niche success (if not in general programming); and some of Bret's explorable explanations ideas [1,2,3] seem to point away from textual programming. [1] http://worrydream.com/#!/ExplorableExplanations http://worrydream.com/#!/ExplorableExplanations [2] http://worrydream.com/#!/InteractiveExplorationOfADynamicalSystem http://worrydream.com/#!/InteractiveExplorationOfADynamicalS... [3] http://worrydream.com/#!/LadderOfAbstraction http://worrydream.com/#!/LadderOfAbstraction