Andrew Leech

> StringIO/BytesIO don't provide support for the desired ring buffer mode of operation Yes of course. The other day I was thinking this PR has a limitation in that a...

> If a mode is supported where the write pointer can overwrite unread data, in that mode the limited thread safety is lost. To be clear the underlying C ringbuffer...

Thanks @glenn20 I had been toying with the idea of adding functions to read/write in chunks rather than byte-at-a-time but figured if I did that I'd need to do some...

I've rebased this with some rework to: * defer to the (somewhat newer) `ringbuf_put_bytes()` / `ringbuf_gett_bytes()` functions. * remove the timeout functions, these were adding bulk without having a real...

I've performed some performance testing to compare the simple byte-by-byte loop to the memcpy based method with the following script. ``` python import micropython try: micropython.ringbuffer except AttributeError: print("SKIP") raise...

> A suitably configured ringbuf ensures that the oldest data is lost. > > An option to specify this mode might be useful. Yeah interesting point, currently it only supports...

Thanks for the review @dpgeorge ! I've addressed your comments in the latest push. I'm getting a fail on the codecov/patch now (81.25% of diff hit (target 98.43%)) due to...

Thanks @dpgeorge I've updated based on your most recent review. Lets see if the overall coverage is improved even more now!

They're helpful suggestions thanks @dpgeorge, all addressed in latest push.

Thanks @dpgeorge that last build failure from the coverage addition has been fixed.