Next message: Tim Ellison OTT: "Labels etc"
Message-ID: <65B141FB11CCD211825700A0C9D609BC01D4D73B@chef.lex.rational.com>
From: "Clemm, Geoff" <gclemm@Rational.Com>
To: ietf-dav-versioning@w3.org
Date: Thu, 2 Mar 2000 09:41:40 -0500
Subject: RE: Extending Target-Selector
From: Tim_Ellison@oti.com [mailto:Tim_Ellison@oti.com]
<geoff>
Depth infinity operations in the stable namespace would return
stable URL's
and depth infinity operations in the user namespace would return
user
namespace URL's.
I assume you were talking about depth infinity operations in the
user
namespace, so I agree with your assumption that they are user URL's
in the
response.
</geoff>
Do we still have target selector 'meta' to denote stable URLs?
I don't recall seeing it in the doc.
This is specified with a "null" workspace selector, i.e.:
Workspace:
<geoff>
As for Target-Selector, I believe our last agreement was to have two
headers, Revision-Selector and Workspace. The Workspace header
affects
version selection for any resource encountered while processing the
request.
The Revision-Selector header only applies to the resource
identified by the
Request-URL, and if that resource is a working resource, revision,
or
versioned resource, it causes the target of the request to be the
specified
revision instead.
</geoff>
Really, the target selector would choose a revision in preference to
a
working resource!
Yup. The Revision-Selector always takes precedence over the Workspace
selector
(but is only applied to the Request-URI resource).
<geoff>
So with depth infinity requests, the Revision-Selector header only
applies
to the Request-URL, and only the Workspace header applies to all the
resources encountered by the request.
</geoff>
Cheers,
Geoff