+WIM files themselves store file and stream names using the UTF16-LE. On
+Windows, wimlib works in UTF-16LE, so conversions are usually necessary and
+there should be no problems with character encodings, except possibly in the XML
+data.
+.PP
+On UNIX-like systems, wimlib works primarily in the locale-dependent multibyte
+encoding, which you are strongly recommended to set to UTF-8 to avoid any
+problems. You can alternatively set the environmental variable
+\fBWIMLIB_IMAGEX_USE_UTF8\fR to force \fB@IMAGEX_PROGNAME@\fR to use UTF-8
+character encoding internally, even if the current locale is not UTF-8
+compatible.
+.SH CASE SENSITIVITY
+By default, the case sensitivity of \fB@IMAGEX_PROGNAME@\fR differs somewhat
+between UNIX-like systems and Windows. WIM images may (but usually do not) have
+multiple files with the same case-insensitive name. Internally, wimlib
+stores filenames as case-sensitive, but on Windows paths
+actually provided by the user for use in a WIM image (e.g. for extracting,
+adding, renaming, or deleting files) will be treated as case-insensitive in
+order to get the "expected" behavior. This differs from the default behavior on
+UNIX-like systems, where such paths will be treated as case-sensitive. Note
+that with case insensitively, a path component may in general be ambiguous due
+to multiple files or directories having the same case-insensitive name. In such
+cases, if there is a file or directory with an exactly matching name, it is
+chosen; otherwise, one of the case-insensitively matching file or directories is
+chosen arbitrarily.