News for hackers. Stuff that matters.
Re:The Next Step for Generics by gorgoiler on Tuesday June 16, 2020 @07:29PM (#23545624)
In my experience I find it very common to write, rewrite and rerewrite code a lot when embarking on a new project. For various reasons, I embark on more new projects than I work on long term ones, which I think is a factor here.

It takes a while for something to settle down enough to say this version is going to be the code for the long run.

At that point, solidity in testing and gradual typing can be added, usually starting with the public interface.

Languages where I’ve had to satisfy all type errors in the various fragments of half brained code I’ve written for version 0.0.85 have always felt a little cumbersome. The compiler’s type checker is a perfectionist looking over my shoulder.

(Go is one of those languages. Ruby is not.)

Re:The Next Step for Generics by trixie_ on Tuesday June 16, 2020 @08:45PM (#23546304)
I'm the opposite. New applications are the ones you refactor the most, and refactoring is an order of magnitude easier/faster when the code is statically typed.

Refactoring any dynamically typed application (regardless of size) is a manual operation, and error prone.

But then again I rarely deal with type errors, when I do it's the compiler looking out for my often 'application ending' mistakes.

Re:The Next Step for Generics by randomdude402 on Wednesday June 17, 2020 @12:11AM (#23547689)
As a person who chooses Python whenever it seems reasonable, I agree.

Go, the language, is simple. Hard to argue with, and sounds great.

The code I have to write though, it's quite simpler to do it in Python more often than not.