FullMetalMeowchemist

Another work around I've found is to use the context manage to open the CloudPath and do the write operation as normal. I'm working with `pyarrow.parquet`, and the [`write_table`](https://arrow.apache.org/docs/python/generated/pyarrow.parquet.write_table.html#pyarrow-parquet-write-table) method...

I'd also be interested for my project. I see #191 is out for this, is this going in any time soon?

If there's anything peculiar, it's that there are mixed files where there are "directories" with more files below, and independent files.  This is not the exact state my bucket...

Sorry for the delay, I did a little bit of debugging today and this is what I found. The following layering/file is what's causing the issue ...