- From: Clemm, Geoff <gclemm@rational.com>
- Date: Fri, 8 Jun 2001 14:13:39 -0400
- To: "DeltaV (E-mail)" <ietf-dav-versioning@w3.org>
From: Stefan Eissing [mailto:stefan.eissing@greenbytes.de]
I see two ways to resolve this issue in the Spec:
1) Top-Down: the spec defines types of resources and for each
type then defines what live properties it holds and what state
it can have.
2) Bottum-Up: The current spec is enhanced by a chapter which
algorithmically defines how to detect the resource type from
the supported-xxx-set. Ideally this would also describe how
future extensions should add to this algorithm for their new
types.
Could the creators of deltaV agree on a way to go forward? Please?
Minimally, I will do #2, by including an enhanced version of Tim's
"property-o-rama" posting (thanks again, Tim!). The discussion
of whether or not to extend DAV:resourcetype as well can go on
in parallel.
Note: one issue with some of the suggested new values for
DAV:resourcetype is that they would breaks the current protocol
semantics that the value of DAV:resourcetype is copied from a
checked-out resource to a version by CHECKIN, and from a version to a
version-controlled resource by UPDATE. Clearly we could fix this up
by talking about what values got added to and deleted from
DAV:resourcetype by these operations, but it is kind of nice and
simple the way it is now.
Cheers,
Geoff
Received on Friday, 8 June 2001 14:14:27 UTC