RE: Guidance of DAV: namespace usage in drafts

I agree that using a custom namespace is the right way to go
for properties defined in individual submissions.
I think it is better for early implementors to have a
"custom else DAV switch", than for the DAV namespace to
become uncontrolled.

Cheers,
Geoff

-----Original Message-----
From: Julian Reschke [mailto:julian.reschke@gmx.de]
Sent: Tuesday, October 23, 2001 4:37 AM
To: ietf-dav-versioning@w3.org
Subject: Guidance of DAV: namespace usage in drafts


Hi,

I think I've come across an issue with the way WebDAV related drafts use the
DAV: namespace. While this mainly applies to individual submissions (like
ours: [1]), it also applies in some way to deltaV and ACL.

In section 1.5 [2], deltaV says:

	Although WebDAV request and response bodies can be extended by
arbitrary
XML elements, which can be ignored by the message recipient, an XML element
in the DAV namespace MUST NOT be used in the request or response body of a
versioning method unless that XML element is explicitly defined in an IETF
RFC.

Which I thing is a good thing. However this leaves us with the question:
which namespace *should* be used in drafts that are *planned* to become an
RFC?

- Putting elements into the DAV: namespace might not be a good idea, *even*
for deltaV or ACL until they reach RFC status. By using a different
namespace, ambiguities about what version of the spec is supported could be
avoided. This would be similar to the W3C practice of using a new namespace
each time a Working Draft changes incompatibily.

- "DAV:experimental" seems to be an option, although this would require that
the registrar of the DAV: URI scheme finally defines the syntax of it
(hint).

- Another approach would be to use a "private" namespace for individual
submissions (probably the cleanest approach).

The reason why I'm raising this right now is that our WebDAV server
(obviously) supports the extensions we have proposed (of which two need
extension elements). The IDs (Internet Drafts) use the DAV: namespace. We
now have the choice of either actually *implementing* the extension using
the DAV: namespace (which doesn't seem to be good style), or to use a custom
namespace (and to bring the IDs in line with that).

Feedback appreciated.

Julian



[1] <http://www.greenbytes.de/tech/webdav/>
[2]
<http://www.greenbytes.de/tech/webdav/draft-ietf-deltav-versioning-20.htm#_T
oc524830510>

Received on Tuesday, 23 October 2001 18:01:12 UTC