3 ms·
I tried to get at this with this paragraph: > This analysis doesn’t get into something that I think is a huge deal: the ecosystem vs the language itself. I onl
by steveklabnik 6y ago
I tried to get at this with this paragraph:
> This analysis doesn’t get into something that I think is a huge deal: the ecosystem vs the language itself. I only track the Rust distribution itself here, but most Rust programmers use many, many ecosystem libraries. When people talk about churn in Rust, do they really mean the ecosystem, not the language? In some sense, doing so is correct: Rust having a small standard library means that you have to use external packages a lot. If those churn a lot, is it any better than churn in the language itself?
The reason I didn't pursue this more is that this is significantly harder to both quantify and get data for. As I said, it is a very important way to think about this problem though!
- neolog 6y agoPackages are part of it. But also even if the packages don't change, the preferred idioms can change as the community figures stuff out.
- steveklabnik 6y agoYeah, it's all sort of the same genre of "stuff that's not, strictly speaking, the language."