News for hackers. Stuff that matters.
Re:38% of bugs at Airbnb could have been prevented by using ... by fosk on Sunday February 10, 2019 @11:38PM (#19132108)
Like everything else, there are pros and cons. It’s not like dynamic languages are worse or better per-se, but they surely make it harder to scale a team or a codebase 10x the original size without introducing what effectively is some sort of typed system (or strict “typed” conventions that the team agrees upon) in place. Which may be done right, or disastrously wrong. A typed language generally doesn’t leave this step up for interpretation later on, since everybody agrees on it since day one.
Re:38% of bugs at Airbnb could have been prevented by using ... by cellularmitosis on Monday February 11, 2019 @06:03AM (#19133553)
As a newbie to clojure, I’m very curious to see what a large codebase looks and feels like. The clojure attitude seems to be that types come at the cost of reduced code reuse, which I tend to agree with. After a decade spent in strongly-typed OO, it is starting to feel like a bunch of sound and fury which seems to end up eventually just getting in my way. I’m definitely open to new ideas in this area, and am looking forward to taking a deeper dive into clojure this year.
Re:38% of bugs at Airbnb could have been prevented by using ... by jondubois on Monday February 11, 2019 @03:27PM (#19137961)
Alternatively, you can break the project up into modules; each of which are maintained my smaller teams. I think that is a cleaner approach.