Dynamic vs Static is a huge debate that I’m not qualified to answer. My personal preference is static because I like to know my mistakes at compile time instead of after running and something weird happens. That goes along with my preference that all variables should be declared at the top of a function.
Yeah I use it too. But when I have to read somebody’s code or my own from a while ago, I prefer everything labeled at the top. That way I can read the top, jump anywhere, and know what is going on without looking at any other lines.
It’s a preference that can be argued like dynamic typing.
I guess you could make a rule of declaring your variables at the top of their scope, be it a class, a function or a code block. That would give clarity without needlessly expanding any scopes.
Well, if there is nof fixed (explicit or implicit) type it’s imposible for the compiler to optimise your code. Also imho programming with typed languages is way easier because your IDE can recognize function argumentd before you compile/run. I tried python and found it baffling how anyone can get any work done with it :D
I used Python almost exclusively before I learned Rust and inevitably became a full time Rust bro, and the answer is “slowly and with a lot of crashes” :P
anyway, as I said in another comment, I’m not talking about static vs dynamic typing, I’m talking about static typing with vs without a compiler that can do type inference. C++'s auto will default to floats if you don’t tell it the type of a number which is pretty brain dead, and there are scenarios where it’s helpful to write out a type that could be inferred for readability/guaranteed correctness’s sake, but apart from that I can’t think why having most of your types be implicit would be bad
why would you not want it to? what circumstance, other than an integer not given an explicit type, could it guess wrong?
Dynamic vs Static is a huge debate that I’m not qualified to answer. My personal preference is static because I like to know my mistakes at compile time instead of after running and something weird happens. That goes along with my preference that all variables should be declared at the top of a function.
https://hackernoon.com/i-finally-understand-static-vs-dynamic-typing-and-you-will-too-ad0c2bd0acc7
I almost upvoted but for that last sentence. Code block scopes are most intuitive, and JavaScript has become a better language since it gained them.
Yeah I use it too. But when I have to read somebody’s code or my own from a while ago, I prefer everything labeled at the top. That way I can read the top, jump anywhere, and know what is going on without looking at any other lines.
It’s a preference that can be argued like dynamic typing.
I guess you could make a rule of declaring your variables at the top of their scope, be it a class, a function or a code block. That would give clarity without needlessly expanding any scopes.
I’m not talking about dynamic vs static though. I’m talking about static typing with or without compiler type inference a la Rust or C++'s
auto
(note that Java making generic parameters optional does not count since that is, in fact, dynamic typing)
I also prefer static typing but I like it when it is implemented like kotlin where type inference is still great, I think dart also works like that
Well, if there is nof fixed (explicit or implicit) type it’s imposible for the compiler to optimise your code. Also imho programming with typed languages is way easier because your IDE can recognize function argumentd before you compile/run. I tried python and found it baffling how anyone can get any work done with it :D
I used Python almost exclusively before I learned Rust and inevitably became a full time Rust bro, and the answer is “slowly and with a lot of crashes” :P
anyway, as I said in another comment, I’m not talking about static vs dynamic typing, I’m talking about static typing with vs without a compiler that can do type inference. C++'s auto will default to floats if you don’t tell it the type of a number which is pretty brain dead, and there are scenarios where it’s helpful to write out a type that could be inferred for readability/guaranteed correctness’s sake, but apart from that I can’t think why having most of your types be implicit would be bad