RE: IE 6.0 cannot handle Japanese yet

Hello,

> -----Original Message-----
> From: Bjoern Hoehrmann [mailto:derhoermi@gmx.net]
> Sent: Thursday, October 25, 2001 11:06 PM
> To: Christian Wolfgang Hujer
> Cc: MURATA Makoto; www-style@w3.org
> Subject: Re: IE 6.0 cannot handle Japanese yet
>
>
> * Christian Wolfgang Hujer wrote:
> >that's wrong.
> >
> >The document of the URL specified is not UTF-8 but probably
> UTF-16 encoded.
> >You have to state any non-UTF-8-Encoding in the XML-Declaration. The
> >xml-declaration of your document is wrong, it has to be <?xml
> version="1.0"
> >encoding="utf-16"?>. So the document is no XML document (it is not
> >well-formed).
>
> Only in absence of higher-level protocol information, you have to
> declare the character encoding in the XML declaration if else than
> UTF-8 or UTF-16, UTF-16 encoded documents require a byte order mark
> but no encoding declaration. Murata-san knows quite well about these
> rules.
sorry, somehow I skipped "UTF-16" when reading the XML spec. Yes, UTF-8 and
UTF-16 do not need an encoding declaration.

>
> >I have tested it, IE 5.5 seems to display it correctly (but it
> shouldn't),
> >but Opera 5.12 and Mozilla 0.9.3 failed to display the document.
> Opera 5.12
> >tried best in displaying [] (unknown char boxes), Mozilla
> complained about
> >the not well-formed document.
>
> Mozilla complains about the byte order mark as invalid character, Opera
> and IE deal just fine with the document, as they should.
>
> >So the CSS document is currently of no interest because the document
> >referencing it is no xml-document, so the xml
> processing-instruction may not
> >be interpreted anyway.
>
> I guess you refer to the xml-stylesheet processing instruction, the xml
> declaration is no processing instruction, even if it looks like one.
Yes, sorry, I meant either the xml-stylesheet processing-instruction or xml
processing-instructions in common (means any PITarget), which one I don't
know any more.

The xml declaration is no processing instruction regarding xml, of course.
(but it is regarding sgml, but this is well-known to us and off-topic)

Greetings

Christian
who is tired now so he won't test if mozilla accepts the document if it
includes the EncodingDecl in its xml declaration.