Frank Tang wrote:
> >  Why do we need unicode paths on the nsIFile?
> 
> File name do contains non ASCII characters. And in general we should use
> wstring/PRUnichar* for all XPCOM interface for any text which are human
> readable. We also want to hide the details of what charset are used in
> the file system from the caller.

While I agree that Unicode is best inside Mozilla (and REQUIRED for
scriptable interfaces), embedding apps may find native code paths easiest to
work with. Why force them to convert to Unicode only to have nsLocalFile
convert them right back?

-Dan Veditz

Reply via email to