The presence of any utf-8 causes markdown parsing to fail with a vague error:
ERROR! The markdown supplied could not be parsed correctly. Did you forget to surround a code snippet with "~~~~"?
The problem only occurs when viewing a markdown file directly, not when it is rendered in a directory view.
Example illustrating the problem, with utf-8 smart quotes (e2 80 99): https://sourceforge.net/u/jnewsome/doctest/ci/405da8825d4b93939f9302421c045e28a73116fa/tree/utf8-fail-demo/README.md
The same file renders fine from the directory view: https://sourceforge.net/u/jnewsome/doctest/ci/405da8825d4b93939f9302421c045e28a73116fa/tree/utf8-fail-demo/