COMPOUND_TEXT versus UTF8_STRING

[Matthias Clasen]

On Wed, 2004-09-22 at 13:58, Roland Mainz wrote:
> Matthias Clasen wrote:
> > 
> > > It is not easy to get rid of COMPOUND_TEXT since many X11 specs have to
> > > be rewritten/updated to do that. And backwards-compatibility to existing
> > > (binary) applications is required, too.
> > >
> > > What about adding an extension to ISO 2022 which defines a new sequence
> > > to mark the following string as UTF-8 - is that possible ? That way
> > > backwards-compatibility to COMPOUND_TEXT is maintained and existing
> > > applications still work without any changes. Is that possible (to be
> > > honestly I don't know much about COMPOUND_TEXT) ?
> > 
> > You don't get rid of a monstrosity by making it even more monstrous...
> 
> The alterantive is to break backwards-compatibility and that will cause
> big interoperability problems. The Linux desktop already has enougth of
> these issues so adding another one will cause even more grief.
> 

COMPOUND_TEXT causes more interoperability problems than it solves.
Introducing new a new variant of COMPOUND_TEXT will not make it better.

Matthias