- From: Geoffrey M. Clemm <geoffrey.clemm@rational.com>
- Date: Thu, 18 Jan 2001 17:38:34 -0500 (EST)
- To: ietf-dav-versioning@w3.org
OK, Greg makes a good point that it is more convenient to do a DELETE than to have to lock some property value and update it. So I'll switch a working collection back to have history resources as members (which is probably better anyway, since it makes a working collection act much more like a checked-out version-controlled collection), but keep the DAV:version-controlled-binding-set for the collection version. I'll try to get this updated draft posted to the web site sometime late tonight. Cheers, Geoff From: Greg Stein <gstein@lyra.org> On Thu, Jan 18, 2001 at 04:22:47PM -0500, Geoffrey M. Clemm wrote: >... > Greg: This means that you would edit the > DAV:version-controlled-binding-set of a working collection to move > existing version-controlled-members into and out of that working > collection. I assume that is fine with you? You would > still add non-version-controlled members to that working collection > if you wanted to create new version controlled resources. >... Hmm... I'm not too sure on this one. I don't do anything with collection versions themselves, but only with working collections. The operations there are: *) DELETE /working/collection/member *) MKCOL /working/collection/newcoll *) PUT /working/collection/newmember *) COPY /some/thing -> /working/collection/newmember The COPY does the whole DAV:precursor-set thing. I never actually "move" something into a working collection, thus editing the binding doesn't apply to me. But! I get a sneaky feeling that the DELETE might not work properly. If the working collection had no members, then I'm guessing that I'd get a 404 on the delete. I'd be rather upset if I needed to PROPFIND/PROPPATCH to do a DELETE operation. (so consider that an objection if I'm correct in my thinking) Cheers, -g -- Greg Stein, http://www.lyra.org/
Received on Thursday, 18 January 2001 17:39:28 UTC