Heapless 1.0

[newAM]

General discussion / notes / thoughts about stabilizing heapless with a 1.0 release.

Checklist

• [ ] Review issues in the v1.0.0 milestone
  • Planned breaking changes are tracked in #494
• [ ] Review Rust API Guidelines Checklist
  • #537
• [ ] Review allowed warnings and clippy lints
  • #532
  • #533
  • #534
  • #536
• [ ] Review non-default clippy lints (-W clippy::pedantic -W clippy::nursery)
  • #542
  • #544
• [ ] Review public enums and structs, should any be marked #[non_exhaustive]?
• [ ] Review explicit panics, unwraps, expects in non-test code
• [ ] Review crate features
• [ ] grep for fixme, todo, ect.
• [ ] Review open issues
• [ ] Review open pull-requests
• [ ] When 1.0 looks ready engage with community, solicit feedback
  • [ ] Announce in rust-embedded WG meeting
  • [ ] Announce in this issue
  • [ ] Anywhere else?

Dependencies

• defmt will reach 1.0 soon: https://ferrous-systems.com/blog/defmt-1-0/
• ufmt may not be 1.0, leave as-is and document it will not be covered by semver?
• hash32 needs a 1.0, https://github.com/rust-embedded-community/hash32/issues/17

Questions

• What should the MSRV policy look like?
  • I think Ferrocene support is important for some users, this is currently at 1.83.0

[rursprung]

see also the discussion on #410 - i think you wanted to push hash32 to 1.0 first (https://github.com/rust-embedded-community/hash32/issues/17)?

[newAM]

Whoops I forgot #410 existed, sorry about that!

[reitermarkus]

Got some more issues:

• [x] Remove Q* type aliases for MpMcQueue, and rename it to just Queue to align with spsc::Queue.
• [x] Don't export indexmap types at the top-level, except indexmap::IndexMap itself. Also, should the module be called index_map instead? Same for indexset module and IndexSet types.
• [x] Align histbuf module and HistoryBuffer type name as history_buf and HistoryBuf?

[pvdrz]

Got some more issues

addressed in https://github.com/rust-embedded/heapless/pull/547