I would welcome this as a compromise, since I currently have to look
through text anyway to entitize other data.
> -----Original Message-----
> From: MURATA Makoto [mailto:firstname.lastname@example.org]
> Sent: Wednesday, December 19, 2001 16:03 PM
> To: Michael Rys
> Cc: email@example.com
> Subject: Re: [xml-dev] Some comments on the 1.1 draft
> From: "Michael Rys" <firstname.lastname@example.org>
> Subject: RE: [xml-dev] Some comments on the 1.1 draft
> Date: Wed, 19 Dec 2001 10:14:59 -0800
> > The normalization of CR, LF and CRLF can be avoided by using
> > character entities. So I still can preserve fidelity.
> You are right.
> How do you feel about allowing character references to C0
> control functions rather than using them directly? I was
> told that such changes are harmful, since bidirectional conversion
> between characters and character references is useful for I18N.
> However, since #&xD; cannot be safely converted to
> such bidirectional conversion is already impossible.