> I think the underlying assumption is more along the lines of: the easy cases are made easier, and the hard cases are made impossible — or rather the concern b

> I think the underlying assumption is more along the lines of: the easy cases are made easier, and the hard cases are made impossible — or rather the concern being that rust hasn’t made the problem easier, it just moved it around. Except in this case the article linked demonstrates that they've made things dramatically easier overall. Data races went from being endemic in a C++ codebase to being in a few low-level and easy-to-locate areas. The easy stuff stayed easy, the bulk of the rest was folded into the easy case, and even the hard cases appear to have been made easier to find, identify, and fix.