I just ran across unintuitive behavior in [String.strip]
When the last U8 preceding the trailing whitespace is a fragment of a utf-8 sequence,
it gets chopped off along with the following whitespace.
Whe the last U8 preceding the end of the string is a fragment of a utf-8 sequence,
that last codepoint gets mangled to its first octet.
This goes against the prinicple of least surprise. Encoding issues notwithstanding,