W3C home > Mailing lists > Public > ietf-dav-versioning@w3.org > January to March 2001

Re: Eliminating the Label header

From: Geoffrey M. Clemm <geoffrey.clemm@rational.com>
Date: Wed, 28 Mar 2001 08:18:21 -0500 (EST)
Message-Id: <200103281318.IAA10955@tantalum.atria.com>
To: ietf-dav-versioning@w3.org

I'd be tempted to not add this yet ... if you are just asking about one
resource, then an extra round trip is not very significant.  If you are
faking baselines with labels, then we are recommending that you just
expose this as baselines (which have a baseline-collection property that
you can run the standard property report against).

If in practice we find this extra round trip to be prohibitively
expensive, we could always extend the property report later (but
looking at the extra syntax now would be a good way to confirm that
we could easily add it later).

Cheers,
Geoff


   Date: Tue, 27 Mar 2001 17:28:09 -0800
   From: Greg Stein <gstein@lyra.org>

   On Tue, Mar 27, 2001 at 01:35:13PM -0500, Clemm, Geoff wrote:
   >    From: Vasta, John [mailto:jvasta@Rational.Com]
   >...
   >    A server may have a much more efficient way to find a version with
   >    a label.  If the find-labeled-version function seems useful, how
   >    about adding a report for it? (The function seems useful to me; I'm
   >    thinking of the number of times I want to compare two versions of
   >    something, where at least one of them is denoted by a label.)
   > 
   > Good point.  We do need to resurrect the "DAV:labeled-version" report
   > (analogous to the DAV:latest-activity-version report), so you can get
   > this information in an efficient standard way.

   Not completely efficient... it adds a round trip.

   With the Label header, I can "redirect through" the VCR and operate against
   the Version Resource. Using the report, I send one request to run the report
   and get the VR, then a second request to operate against the VR.

   What would be really neat is a way to specify a label within an
   expand-property report. For my particular use-case, that would collapse a
   whole sequence of PROPFINDs into a single report. Is there any interest in
   allowing this? (i.e. should I propose a syntax?)

   Cheers,
   -g

   -- 
   Greg Stein, http://www.lyra.org/
Received on Wednesday, 28 March 2001 08:20:06 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 13:57:41 GMT