- From: William F. Hammond <hammond@csc.albany.edu>
- Date: Wed, 1 Aug 2001 14:07:15 -0400 (EDT)
- To: www-html@w3.org, www-talk@w3.org
Sorry to find the need to belabor this.
Masayasu Ishikawa <mimasa@w3.org> writes to www-html:
> [1] http://www.w3.org/TR/ruby
A well-written W3C recommendation.
As things are by default (in a platform OS that correctly does not
arrogate to itself knowledge of the meaning of a "suffix" found in a
URI), its normative appendix A cannot be read in Amaya's window.
I'm not saying that "application/xml-dtd" is wrong, but it seems to me
in this case not to be the best choice by the content provider.
I am inclined to construe the use of the word "should" in RFC 3023,
section 3, para 2, to be for mandating application/xml-dtd over either
text/xml or application/xml but not to preclude fallbacks by a content
provider to text/plain or application/octet-stream, as deemed
appropriate by the content provider.
Should RFC 3023 be interpreted otherwise? Might the authors of RFC
3023 be inclined to clarify this?
Please note that if a user configures a user agent to call a
text/plain reader for application/xml-dtd, this will preclude other
use of that content type by the user agent. That other use in other
contexts might, in fact, be the intent of the authors of RFC 3023.
-- Bill
Received on Wednesday, 1 August 2001 14:07:25 UTC