[MPlayer-dev-eng] [RFC] Use utf-8 for all source files?

[Rich Felker]

On Mon, Mar 12, 2007 at 06:22:28PM +0100, Diego Biurrun wrote:
> On Mon, Mar 12, 2007 at 06:12:35PM +0200, Uoti Urpala wrote:
> > On Mon, 2007-03-12 at 16:53 +0100, Diego Biurrun wrote:
> > > On Wed, Feb 28, 2007 at 06:52:22PM +0800, Zuxy Meng wrote:
> > > > 2007/2/28, Diego Biurrun <diego at biurrun.de>:
> > > > >What source files are affected?  The idea is really that all sources
> > > > >should be ASCII and all the rest UTF-8.  I think this is the decision we
> > > > >came to when discussing this some time ago because some (legacy?)
> > > > >compilers might possibly choke on UTF-8.  latin-1 is bad, this needs to
> > > > >be fixed in any case.
> > 
> > I see no problem with using some utf-8 in comments for author names and
> > such. Compilers shouldn't be a problem; if they haven't choked on
> > latin-1 they should not choke on utf-8 (as latin-1 allows MORE byte
> > sequences than valid utf-8). And compilers which have problems with
> > non-ascii could not compile the current translation files.
> 
> I don't remember the details, this was discussed a long time ago.  Does
> anybody remember the reasons?  Rich?

I don't remember any such issue. I'm all in favor of using UTF-8 in
author names where needed. The output for these strings would be
broken (displaying 2-3 codepage chars per UTF-8 sequence) if someone
built MPlayer with a legacy charset as the internal encoding (still
possible IIRC), but that's much less bad than the old situation with
latin1, where conversion would encounter "illegal sequence" on these
strings!

Rich