2 ms·
Very cool. I like this kind of stuff. One of the links there is about Dirac's notation. When I was doing Quantum Computing in school, professor started us wit
by rdtsc 10y ago
Very cool. I like this kind of stuff.
One of the links there is about Dirac's notation. When I was doing Quantum Computing in school, professor started us with matrices and vectors. We were familiar with those and we could multiply them and do basic operations.
Then one homework they assigned us a more complex quantum circuit to solve. And it was quite painful because matrices had gotten bigger. However there was a method to his madness, because after they introduced Dirac notation. The power of a better notation because obvious to us. It is like programming in C doing some text procecessing or some command line scripts, then someone showing you Python.
Programming is notation in a way. So knowing different paradigms and frameworks is important as some might allow for a very concise representation of the problem.
There was a link here on HN about using Prolog to represent and work with a type system, and it was just very elegant how it worked.
Object oriented design, (which we like to hate today it seems), was seen as a powerful way to represent entities so it became useful. Then design patterns came about, that is also a notation ("Wow there Jim, looks like you've go a singleton, wrapping a factory with an observer flyweight..."). I joke, but it did (still does?) help talk about architecture.
Functional programming is the same way -- learning it teaches you to think in a different way. When state is mutated it is not just s.add_foo(foo) but looks like new_s = add_foo(s, foo). That seems trivial and silly, but it changes how you think about data and state. Or I remember when closured "clicked" for me and I realized how similar they are to class instances when they capture their environment.
Or say thinking how your language approaches concurrency -- is it based on actors and messages, threads, callbacks (maybe with a reactor loop), promises, transactional memory, and so on. Those are notations to solve "concurrent things" in a way. Some might be a lot more efficient at representing the problem than others. But if you never know about it or used it before, it will be hard to convince yourself of it usefullness.
- GavinMcG 10y agoIf you or someone could track down the Prolog/Type System post, I'd be interested!
- Avshalom 10y agohttps://news.ycombinator.com/item?id=12108041 https://news.ycombinator.com/item?id=12108041
- ci5er 10y agoI have to imagine that attempting long-division or creating log tables or transcendental function tables would have been quite the formidable task in roman numerals. Notation matters.