To answer #1uwekel wrote:1. Why does it perfectly work under Linux (i guess because Windows codepage)?
2. How can i get the same result on Windows as on Linux?
Probably because Linux ditched Ascii ages ago in favour for Unicode. On Linux "Ascii" is really UTF8.
On Windows Ascii is 7-bit. 8 bit strings on Windows are not UTF-8, they are instead codepages and each language tens to have their own. So German Windows will probably have a German codepage (or possibly Latin-1 / ISO 8859-1).
As to #2
You may have to look into the Codepage/Multibyte Windows APIs. You will need to convert from the system codepage to unicode.
Tip!
If possible, make sure the server/machine you are fetching the data from is sending you UTF-8 strings/text instead of Ascii. That way you can use Peeks with the #PB_UTF8 flag and it will remain correct regardless of system language or system default codepages.
When writing/sending strings/text you should make sure that the text is also UTF-8 (use PokeS and #PB_UTF8 flag for that).