Stream: t-libs/wg-allocators

Topic: Collections beyond box, allocation fallibility


John Ericson (May 03 2019 at 17:04, on Zulip):

Once Box is done, I'll rebase https://github.com/rust-lang/rust/pull/52420 . The interesting thing is that this also solves the issue of fallible allocation by giving Alloc and associated error type.

Tim Diekmann (May 03 2019 at 17:15, on Zulip):

Oh, I completely forgot that associating an error type to Alloc is still a thing.

Simon Sapin (May 03 2019 at 17:49, on Zulip):

I disagree that an associated error type “solves” infallible allocation. Returning Result<T, !> or Result<T, Infallible> is not the same as returning T.

Simon Sapin (May 03 2019 at 17:50, on Zulip):

It’s a thing to the extent that we want to make it a thing. A new issue in the repo would be a good place to discuss use cases.

Last update: Nov 15 2019 at 10:10UTC