If there were no classes, subclasses, objects, complex types and all the abstraction clusterfucks, and you just had to deal with simple things, like vars/consts and functions, how would that affect your programming?
Wouldn't you be able to do your work? If you don't need them (in theory nobody does - the can just copy/paste similar code to emulate objects or make multiple declaration instead of using complex types), you could just skip them all together or bypass them. If the cpu can do everything by 10-15 math and logical functions, like comparing, jumping, adding, moving data, so can we at a much higher level (but still not too high to create clusterfucks).
Wouldn't you be able to do your work? If you don't need them (in theory nobody does - the can just copy/paste similar code to emulate objects or make multiple declaration instead of using complex types), you could just skip them all together or bypass them. If the cpu can do everything by 10-15 math and logical functions, like comparing, jumping, adding, moving data, so can we at a much higher level (but still not too high to create clusterfucks).
OMG, I will let smooth handle this one or just ignore it. I'm not up for teaching you why your conceptualization is extremely naive.
As I said, I'm not a programmer and thus my use cases for "complex" language features have always tended to zero. And even if I were, I'm not sure I'd prefer the more complex language, just for the sake of it. For example is ad-hoc polymorphism a necessity for my use case or can I do the same by emulating polymorphism and breaking it down? If I can break it down, I will - instead of, say, going to a more complex language. At least that's my theory of what I expect I'd do if faced with a problem.
In any case I think you said one of the applications you created was in assembly. Obviously the language problems didn't affect you - the sky was the limit (=the hardware limitations were the only limit). And I doubt the code of the 80s was too "evolved" in terms of complexity anyway. I've seen sources from the 70s and 80s - even for things like compilers and OSes... they are full of if/then/else, loops, functions and that's pretty much it. It's like a more "readable" version of the underlying assembly expressed in c, pascal or something similar. And you know what? The software back then was way more reliable than today's. Then you had an OS or an app, and it wasn't *supposed* to be followed by 1500 patches. It was supposed to work out of the box, as intended. I don't know if it was because it was broken down to simple uses of language or if they tested the code 100 times more, or that the uses of the code were more limited and thus the code much simpler but that was the case. Something went wrong since then.
Quote
Start reading from here: https://users.rust-lang.org/t/high-order-function-with-type-parameter/3112/101
I'm reading it, but I don't understand everything.
Quote
P.S. with first-class functions and closures, you can model most of the same semantics, but it gets hairy.
Hmm...
