Hello, On Mon, Jan 17, 2011 at 9:33 AM, Sukender <[email protected]> wrote: > I would then recommend to keep OSG use "narrow" std::string, and have a flag > somewhere (Some singleton? Registry? ReaderWriter::Options?) which says if > filenames use a (8-bit) code page or UTF8. Having "wide" UTF16 overloads may > be possible too (even if conversion functions are enough). Thoughts?
I am jumping into the middle of the discussion here, but something like this is normally handled by locale settings on every OS. There is little point setting this separately, the OS determines this. E.g. Windows uses UCS-2 for filenames on NTFS and ASCII on FAT, almost every Unix is able to work with UTF-8 encoding for filenames today. Whether the filenames are ASCII + codepage or UTF-8 is determined by the LC_xxx variables on Unix, not sure how the locale is set on Windows, but there is probably some registry key for this. jan _______________________________________________ osg-submissions mailing list [email protected] http://lists.openscenegraph.org/listinfo.cgi/osg-submissions-openscenegraph.org
