W3C home > Mailing lists > Public > ietf-dav-versioning@w3.org > January to March 2002

RE: Subversion/WebDAV

From: Clemm, Geoff <gclemm@rational.com>
Date: Tue, 12 Feb 2002 07:15:44 -0500
Message-ID: <3906C56A7BD1F54593344C05BD1374B105CE0B4A@SUS-MA1IT01>
To: Julian Reschke <julian.reschke@gmx.de>, Greg Stein <gstein@lyra.org>
Cc: "DeltaV (E-mail)" <ietf-dav-versioning@w3.org>
   From: Julian Reschke [mailto:julian.reschke@gmx.de]

   > From: Greg Stein [mailto:gstein@lyra.org]
   >
   > On Mon, Feb 11, 2002 at 11:47:51AM +0100, Julian Reschke wrote:

   > > Subversion reports the DAV:version-name property for VCRs. This
   > > breaks our WebDAV adapter which assumes that the presence of
   > > this property indicates that this is a version resource.

   > Hmm. It isn't supposed to do that? I thought a VCR was supposed
   > to have a copy of all the version's properties. Is
   > DAV:version-name excluded from that? Am I mis-remembering
   > something?

   Good question. Maybe this needs to be clarified in DeltaV
   (Geoff?). As I said, we are using the presence of this property to
   detect that something is a version resource.

A VCR should have a copy of all the version's *dead* properties,
but definitely should not have a copy of the version's live
properties (since, as in Julian's case, clients will use the
presence and absence of these live properties to determine
what kind of resource is there).

In particular, you will note in 22.5 that a VCR does not have
a DAV:version-name, while in 22.6, it states that a version does.

Cheers,
Geoff
Received on Tuesday, 12 February 2002 07:16:50 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 13:57:43 GMT