- From: Jim Whitehead <ejw@cse.ucsc.edu>
- Date: Mon, 19 Nov 2001 17:10:27 -0800
- To: <acl@webdav.org>, "WebDAV" <w3c-dist-auth@w3.org>
It seems reasonable to me to add this text. I'd also go further, and explicitly note that validating requests and responses is generally a bad idea. IETF protocols should be strict in what they send, liberal in what they accept. Strict checking of a DAV message for validity is more stringent than is necessary to interpret the meaning of the message. It has been my experience that implementations that do require strict validity tend to be much less interoperable, since they tend to reject XML that most other implementations accept without any problem. This is why the DAV spec. has never required anything more than well formedness. - Jim > Below is an attempt to clarify the role of the DTD fragments: > > 19 APPENDICIES > > 19.1 WebDAV XML Document Type Definition Addendum > > All XML elements defined in this Document Type Definition (DTD) belong to > the XML namespace "DAV:"DAV namespace. This DTD should be viewed as an > addendum to the DTD provided in [RFC2518], section 23.1. > > Note that WebDAV messages must not be validated using the DTD, as > WebDAV is based on XML namespaces, and the special WebDAV "element > ignore" rule ([RFC2518], section 23.3.2) applies. > The following transformations need to be applied to a WebDAV > message before it can be validated using the DTD: > - removal of all elements and attributes not defined in > RFC2518 or this specification, > - transformation of all remaining elements into elements in > no namespace. >
Received on Monday, 19 November 2001 20:10:45 UTC