[RFC] blocking/serial: make method naming consistent

[Dirbaio]

blocking::serial::Write's methods are named bwrite_all and bflush.

• the b prefix indicates blocking (I guess?). However no other blocking trait method has this prefix.
• the _all suffix indicates it writes a slice instead of a single word (I guess?). However, no other trait method that writes a slice has this suffix.

To make them consistent, this PR renames them to write and flush.

[rust-highfive]

r? @eldruin

(rust-highfive has picked a reviewer for you, use r? to override)

[eldruin]

For completeness, as I highlighted in #281, different method names ensure unequivocal selection between different traits when they are all pulled in by the prelude.

• dev.read() -> nb::serial::Read::read()
• dev.write() -> nb::serial::Write::write()
• dev.flush() -> nb::serial::Write::flush()
• dev.bwrite_all() -> blocking::serial::Write::bwrite_all()
• dev.bflush() -> blocking::serial::Write::bflush()

The method names can certainly be improved, though. Removing the prelude is also an option.
With the same method names users need to either stop using the prelude or turbofish calls.

[ryankurte]

the blocking/non-blocking delineation makes more sense than the send/write difference to me, i suppose we could also split the prelude for blocking/non-blocking/futures?

[Dirbaio]

the blocking/non-blocking delineation makes more sense than the send/write difference to me,

Agreed, but then all blocking methods should have the b prefix, not just UART. Or all the nb methods should have the nb_ prefix.

i suppose we could also split the prelude for blocking/non-blocking/futures?

That won't work nicely because people end up using both blocking and nb traits in the same mod. Especially since some traits are only available in blocking flavor and others only in nonblocking flavor.

Nobody uses the same trait in both blocking and nonblocking flavors at the same time though. IMO naming blocking and nonblocking the same is not that bad.

[eldruin]

Looks good to me, thanks!

[Dirbaio]

ping @ryankurte @therealprof is there anything missing to get this in? also for #282

[therealprof]

LGTM, bors r+

[therealprof]

bors r-

[therealprof]

Should have a CHANGELOG.md entry.

[therealprof]

LGTM now, thanks.

bors r+

[bors]

Build succeeded:

• ci-linux (1.35.0, x86_64-unknown-linux-gnu)
• ci-linux (stable, thumbv6m-none-eabi)
• ci-linux (stable, thumbv7m-none-eabi)
• ci-linux (stable, x86_64-unknown-linux-gnu)
• fmt