- From: David Dorward <david@dorward.me.uk>
- Date: Fri, 20 Jul 2007 16:56:16 +0100
- To: www-validator Community <www-validator@w3.org>
On 20 Jul 2007, at 16:46, Sierk Bornemann wrote: >> If there is no XML declaration, then XML parsers, IIRC, are >> supposed to assume XML 1.0 and UTF-8 or -16 (depending on the BOM, >> if present). > > If an XML parser is to be applied at all and not a SGML parser, > depending on which MIME type is served... > But that's another hot discussion... Entirely. Although if a text/html content-type is used, then Appendix C recommends you don't use an XML declaration at all. >> What's the point of adding the XML declaration under any >> circumstances short of changing output language or encoding? > > For instance to be as compliant and close to the XHTML 1.0 > Recommendation as possible? Which uses the weak term "strongly encouraged", which doesn't make it a requirement, and doesn't use RFC language such as "SHOULD". > Remember: Serving a XHTML 1.0 document as "text/html" is a "MAY" > whereas serving it as "application/xhtml+xml" (and so serving it as > XML) is a "SHOULD". > "SHOULD" has a stronger meaning than "MAY" (in contrast: "SHOULD > NOT" has a weaker meaning than "MUST NOT"). This is the content-type issue again. -- David Dorward http://dorward.me.uk/ http://blog.dorward.me.uk/
Received on Friday, 20 July 2007 15:56:49 UTC