I know the potential pitfalls of slices, and how to avoid them. I’ve read about the details, etc. But, personally, I think that fixing slices should be #1 on the list for go 2.0. The fact that I have to study it and work to not get tripped up on slices is the biggest flaw of the language, IMHO. Fixing slices would be worth the breaking change.

BTW, I was convinced at GopherCon 2019 that incorporating generics is a good thing, but that’s not gonna require a 2.0, it sounds like. Perhaps there’s a path of implementing generics (1.x), replacing maps and slices with new explicit generic solutions (1.x+n), then we can move to just dropping the current map & slice (2.0).