W3C home > Mailing lists > Public > w3c-dist-auth@w3.org > October to December 2001

RE: RFC2518 (WebDAV) / RFC2396 (URI) inconsistency

From: Julian Reschke <julian.reschke@gmx.de>
Date: Sun, 25 Nov 2001 23:20:48 +0100
To: "Lisa Dusseault" <lisa@xythos.com>, <w3c-dist-auth@w3.org>
Message-ID: <JIEGINCHMLABHJBIGKBCMEIADIAA.julian.reschke@gmx.de>
> From: w3c-dist-auth-request@w3.org
> [mailto:w3c-dist-auth-request@w3.org]On Behalf Of Lisa Dusseault
> Sent: Sunday, November 25, 2001 7:55 AM
> To: w3c-dist-auth@w3.org
> Subject: RE: RFC2518 (WebDAV) / RFC2396 (URI) inconsistency
>
>
> The pragmatic arguments: I don't think it's really feasible to change the
> bits on the wire now.  Obviously, it hasn't been a problem in the

We can't change them now, but we can define a migration path to a format
that conforms to the underlying specs.

> past, and
> it hasn't harmed interoperability.  I've never heard of XML parsers that
> object to that namespace, and I'm not clear why the single tool that does
> object to that namespace is required for DAV servers/clients.

It isn't.

But it *is* an example of a newly written piece of code that rejects the
namespace name, and we have found out that it's correct in doing that. So we
have to consider the fact that this may apply to newly written XML
processors as well.

> Changing the namespace DAV uses in all its bodies would be a tremendous
> change -- and I don't see what the need is for such a major change.

Because the standard as it's written today isn't well-defined? It says that
you must use an XML-NS aware processor to parse the bodies, yet the bodies
as defined by the spec do not conform to XML 1.0 (+ namespaces).

> Besides, even if we tried to "fix" RFC2518 on this issue, that's not the
> only place where it crops up.  I know of other namespaces used in shipped
> software products where the namespaces look like a string of characters
> ending in ':'.   Changing the "DAV:" prefix won't fix those, whereas a

Well. Obviously these need to be fixed as well.

> loosening of either the XML namespaces requirements or the URI
> requirements
> would.
>
> The theoretical argument I haven't heard yet:  If the definition of this
> type of URI is "<scheme>:<scheme-specific-part>", why can't the
> scheme-specific part string be the null string -- as long as that's legal
> for the declared scheme?

I was wondering about this as well.

Fact is the grammar doesn't allow it, and it doesn't seem that anybody is
going to touch RFC2396 "just" for WebDAV.

Julian
Received on Sunday, 25 November 2001 17:21:25 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 2 June 2009 18:43:59 GMT