re: Determination of Encoding

Gavin Nicol <gtn@eps.inso.com> wrote:
> Precisely why I way that we must rely on HTTP header. I'm starting to
> think that Rick's proposal of requiring servers to remove the PI
> is a good idea.

If servers remove it, it's better not to have it in the first place.

> I have argued for treating the PI as a header a number of times, or
> using an alternate header syntax.

I agree, and your arguments were sound, but not I think understood.
Using MIME headers at the start of an XML file -- but _not_ at the
start of the corresponding SGML entity -- is perfectly Kosher SGML,
and would be a lot easier to deal with.

Another alternative would have been to have wrapped the XML document
in a simple XML envelope; with namespaces, this no longer pollutes the
document's namespace...

<XML RMD="no" XMLREV="2">
  <HEAD>
    <META NAME=CHARSET HTTP-EQUIV="..." CONTENT="iso8859-27">
    other stuff here.
    Note that this is HTML-compatible, for servers such as CERN
    that read http-equiv and put it out in headers...
  </HEAD>
  <INSTANCE>
    Your document goes here
  </INSTANCE>
</XML>

But I don't think that will happen.

Even so, I still think that
    Encoding: iso8859-27
    Content-type: text/xml

    Your document goes here
is sensible, and has the advantage that if the server strips the
headers and puts them only in the http header, they needn't get lost,
but if they are passed through, that's good too.

This could be done even in the presence of the <?XML...?> thing.

Lee

Received on Thursday, 19 June 1997 21:05:38 UTC