r? @nikomatsakis

(rust_highfive has picked a reviewer for you, use r? to override)

This makes tail() inconsistent with tail_mut(), init(), and init_mut().

I'm working on a separate PR that implements my proposal, of replacing tail() / init() with pop_first() / pop_last().

My PR has been submitted as #24184.

Hmm. @kballard's PR is a bit more ambitious, so they're not directly comparable; but at minimum I agree that having tail be inconsistent with the other fns doesn't seem quite right!

So I'm going to leave this PR open as an alternative to #24184. I figure we might decide to take this PR, #24184, or neither. But let's move the conversation to #24141 so it can be consolidated.

☔ The latest upstream changes (presumably #24674) made this pull request unmergeable. Please resolve the merge conflicts.

I think I'm going to close this due to rust-lang/rfcs#1058 -- let's see how that pans out, but that approach in some form seems to be popular.

Thanks @GuillaumeGomez for the PR in any case!