RE: need clarification on 2.13 (Normalization Checking)

> From: John Cowan [mailto:jcowan@reutershealth.com]
> Sent: Sunday, May 26, 2002 7:08 PM
> To: Julian Reschke
> Cc: www-xml-blueberry-comments@w3.org
> Subject: Re: need clarification on 2.13 (Normalization Checking)
>
>
> Julian Reschke scripsit:
>
> > As far as I understand the requirement, any XML 1.0 document that uses
> > Unicode composing characters would become non-wellformed in 1.1, right?
>
> Only if they are used where precomposed characters are already available;
> thus A followed by COMBINING ACUTE is non-WF because A WITH ACUTE
> is available.

OK, this is the case here.

> > If this is the case, protocols that currently use XML 1.0 for
> marshalling
> > will either
> >
> > - have to stay with XML 1.0 or
> >
> > - ensure that composing characters do not appear in protocol messages.
>
> XML is not meant to be a dumping ground for arbitrary sequences of
> Unicode characters without regard to sense.

I didn't mean to say that. It's just that I think that this change will
break more than the editors seem to think, and that those who are affected
are not aware of that. I personally believe that Unicode normalization would
be a good thing.