Skip to content

Latest commit

 

History

History
4 lines (3 loc) · 2.59 KB

Studio for the 21st cen chip.md

File metadata and controls

4 lines (3 loc) · 2.59 KB

I was thining about the programming language for this thing. And as far as I can see the best language is visual programming language. Not only the forms would be visual everything will be visual. Loops, call to functions, the classes themselves everything is visual, and immedietly compilees into FPGA code circuitry. This visual language is not complex to define and I'm thinking in it all the time to design CGOL creatures. People are just thinking about those things in different level still, but the basic idea is similiar. There is nothing in computer languages that can't be represented visually or improved by representing it visually. We only need to make the connection to the classes and function on the chip but this is fine. We need those classes that make sense to a lot of people. Customizable structures that people can enter and calculate very fast and use the classes in the most meaningful way for maximum people. So it would be like geometry. It will be built on simple operations like "and"/"or"/"xor" if and change state. Or we will find better representations of the ideas based on visual ways we think about the stuff in CA.

So there would be no regular assembly but some sort of two or three dimenssional assembly that will allow to code programable arrays in sophisticated manner and design classes and functions. And the classes and ideas will be immedietly represented on the chip. Similiar to the way we code now, only a lot of standard libraries will be on the chip as well. And the outer syntax or visual language can be also done in several layers. We need to think visually about the programming - we should stop thinking command line style. The visual thinking of CA is the correct thinking about programming. It's programming - not the textual thing. Text is bad - it reduces information, I want to draw and show pictures. I want to have class Cat and a simulation of a Cat and a physical cat to be on the same line in some coordination system. And I want to say the simulation of the cat is not real enough, who's fault is that? And in more abstract scenarious I want 3d dimentional cubes which I can move select find and connect between them. Just like working shop with many parts and I move the parts and connet between them and build a new library of new parts etc. And some parts I sell on the market, and the market is this chip which sees how many people use what function, and it prioritizes the function with more usage and hard prints it on the chip. And the hard printing is just improve in performance. And people post their functions and help each other to compete and make the best useful function.