Alex Villacís Lasso <a_villacis(a)palosanto.com> writes:
I was rather hoping for an explanation of which is the "correct" behavior for an UTF-8 locale: 1) Open File Dialog returns an UTF-8 encoded string (visible to the application, current behavior), and open-file functions expect UTF-8 2) Open File Dialog returns locale-encoded string (even in the UTF-8 locale), and open-file functions expect locale-encoding (as they do now)
2) is the right behavior. The A functions always return strings in the Ansi codepage, not in the Unix one. There is no Windows locale that uses UTF-8 as Ansi codepage, so if a UTF-8 string is returned to the application that's a bug.
Your comment strongly suggests (2) is the correct approach, but what happens in East Asian locales? Or am I just demonstrating a lack of knowledge on how non-UTF8 encodings work?
Asian locales will use one of the double-byte codepages. -- Alexandre Julliard julliard(a)winehq.org