- From: William F. Hammond <hammond@csc.albany.edu>
- Date: 21 Oct 2001 13:08:09 -0400
- To: www-html@w3.org, iesg@ietf.org
Bjoern Hoehrmann <derhoermi@gmx.net> writes in www-html@w3.org: > * Masayasu Ishikawa wrote: > >Bjoern Hoehrmann <derhoermi@gmx.net> wrote: > >> Since Appendix C ("HTML Compatibility Guidelines") defines mandatory > >> requirements for text/html delivery of XHTML documents ([1]) > >No. > >> [1] mandatory through normative section 5.1, I believe; > > > >Section 5.1 merely says "may". > > Delivering XHTML 1.0 as text/html is a "may", following the guidelines > isn't. Appendix C is labeled _informative_. It "summarizes design guidelines for authors who wish their (X)HTML documents to render" in old user agents. The mime labeling decision belongs to the content provider as agent for the author. Is anyone saying that Amaya is wrong in handling (X)HTML extended by MathML whether served as "text/html" or as "text/xml"? -- Bill P.S. Those who favor "application/xhtml+xml" need to think ahead on the issue of how that information, which would be part of a user-owned data structure, would survive delivery to an http server running on a primtive platform where content types are determined by the 1980's-style canonization of file system suffices. P.P.S. Remember that the name of the W3C document type for (X)HTML is "html". Don't confuse that with its formal public identifier or its XML namespace.
Received on Sunday, 21 October 2001 13:08:15 UTC