Hello Krasimir, Friday, February 03, 2006, 2:03:20 PM, you wrote: KA> * Do you think that we have to use unboxed arrays of Word8/Word16 KA> for file path? utf8-encoded FastPackedString is much more appropriate and about unicode filenames in windows - i have a library what supports such filenames in NT/2000/XP systems. i plan to incorporate this functionality in Standard Hierachical Libraries AS> The task: Remove all files in a directory recursively. AS> The problem: In case the current encoding is UTF-8, filenames stored in AS> a different locale can comprise illegal UTF-8 sequences and are AS> therefore not representable as FilePath which is a Unicode string. Even AS> if the resulting Unicode sting is not 'error ".."', it is impossible to AS> call 'delete' on that file name, since fromUTF8 . toUTF8 cannot be the AS> identity function if the UTF8 byte sequence is illegal. AS> The solution: FilePath must be an abstract data type that is a sequence AS> of bytes. Programmers should only convert these to Unicode for AS> displaying them and otherwise treat them as opaque entities. In case of AS> invalid UTF-8 strings, the corresponding String will have an "invalid AS> unicode code character" substituted. but even my library can't really solve all problems. win95-based Oses don't support unicode APIs, afaik -- Best regards, Bulat mailto:bulatz@HotPOP.com