Upper/lower case is locale language sensitive. One specific example, the uppercase for "i" varies depending on whether you have a Turkish locale or not. String

Upper/lower case is locale language sensitive. One specific example, the uppercase for "i" varies depending on whether you have a Turkish locale or not. String.toLowerCase() in Java has exceptions coded in for the "tr", "az", and "lt" locale languages. Some of these languages may cause the length of the string to change when upcased or downcased, too. Additionally, given a utf8 string, OSX will translate it into another, possibly different utf8 string before using it as a filename (the NFD normalization that Linus mentioned).