Re: GSString problem

[Richard Frith-Macdonald]

On 5 Jul 2004, at 15:51, Alexander Malmberg wrote:

> Richard Frith-Macdonald wrote:
> > On 5 Jul 2004, at 15:22, Alexander Malmberg wrote:
> [snip]
> > > Do we really ever want to set the c-string encoding to
> > > NSASCIIStringEncoding instead of NSISOLatin1StringEncoding? 
> > > (Especially
> > > since parts of Unicode.m is pretending that ascii==iso-8859-1 
> > > anyway.)
> >
> > I guess we do if the user wants to ...
> > ie if GNUSTEP_STRING_ENCODING specifies ASCII
>
> Fair enough, but can you think of any reason why a user would ever do
> that? :)

I try not to second guess people.  I don't think *I* am ever going to 
do that,
but perhaps someone might want to use strict ascii to enforce some sort
of portability rule they have invented.

> > However, I think it's probably sensible to use ISO-8859-1 if
> > GNUSTEP_STRING_ENCODING
> > is not defined and nl_langinfo(CODESET) says ASCII
>
> Agreed.
>
> > I wasn't aware that Unicode.m ever pretended ascii==iso-8859-1 ... I
> > don't think it should.
>
> GSToUnicode silently accepts input bytes >=128 for
> NSASCIIStringEncoding, effectively treating them as iso-8859-1. Not 
> sure
> if there are any other places left.
>
> Currently, this has the advantage that stray 8-bit values in eg.
> filenames won't cause horrible problems by default. If iso-8859-1 is
> made the default again, I think it should be fixed.

OK ... I think that's in CVS properly now.