Can please someone explain the following problem: I have a Windows 95 box, Czech edition (it uses eastern-european windows code page cp 1250 for character encoding) named flex with a share called c . On c, there is a file which contain some czech characters in its name. I have another box with a fresh fedora linux instalation and a freshly compiled samba 3.0.2. UTF-8 is used for character encoding. I can read files on flex by: smbclient \\\\flex\\c password or even transfer them. I don't have any smb.conf, just use smbclient. But the czech characters are wrong. They look like the smbclient assumed western character encoding. After reading docs, I assumed that smbclient >3.0 will negotiate the code pages with clients on the fly. What is wrong? Can I set manually clients code page on the smbclient command line ? Best regards -- Ladislav Sl?de?ek informatik Nemocnice P?erov Dvo??kova 75 751 52 P?erov