On Tue, Feb 19, 2008 at 09:41:03PM +0100, Joerg Schilling wrote: > Alan Coopersmith <Alan.Coopersmith at sun.com> wrote: > > Nicolas Williams wrote: > > > At the very least we need to support non-UTF-8 locale selections by the > > > user without having to update this file, and if that's done then why > > > bother with this system-wide setting? > > > > Solaris is phasing out non-UTF-8 locales, and it's getting harder and harder > > to login to one.
I am aware. > UTF-8 does not solve all problems and it even introduces other problems. I agree, but this comment is not relevant to the issue I brought up. [OT: That said, I think UTF-8 is the only workable solution for filesystem object names. Desktops and GUIs should be able to deal with that, and other charsets should be usable for other content where limitations of Unicode may be unacceptable. If you wish to have a discussion about *this* then let's continue it elsewhere (i.e., not on lsarc-ext).] Alan's comment is relevant because he's saying that we're effectively saying that we're going towards using only UTF-8 for filesystem names (which I think is the right approach), so we should use UTF-8 as the charset for these well-known names even when the user's locale is not a UTF-8 locale. I think that's good enough for me. Nico --
