Description of problem:
When using UTF8, underscores should come before letters.
# printenv LANG
# LANG=C ls
This is not to do with UTF-8, but with en_US, and the collation
behaviour you see is correct and mandated by POSIX. See also bug #76328.
*** This bug has been marked as a duplicate of 103539 ***
Changed to 'CLOSED' state since 'RESOLVED' has been deprecated.