- From: Wilfredo Sánchez Vega <wsanchez@wsanchez.net>
- Date: Fri, 23 May 2008 09:54:50 -0700
- To: Julian Reschke <julian.reschke@gmx.de>
- Cc: Cyrus Daboo <cyrus@daboo.name>, Geoffrey M Clemm <geoffrey.clemm@us.ibm.com>, acl@webdav.org, Konstantin Breu <Konstantin.Breu@gmx.net>, 'WebDAV' <w3c-dist-auth@w3.org>
On May 23, 2008, at 9:20 AM, Julian Reschke wrote: >> Now, WebDAV ACL makes it clear (first paragraph of Section 2) that >> one of DAV:principal-URL or DAV:alternate-URI-set MUST contain a >> URL to a principal resource (http/https). So DAV:current-user- >> principal-resource is just that value. DAV:principal-URL can >> therefore be any URI scheme without affecting that. > > OK, so the proposal for DAV:current-user-principal-URL^H^H^Hresource > would work independently of the other issue, right? Ah, yes. Though if you want to use the DAV:current-user-principal- URL^H^H^Hresource property to fine the URL you can use for yourself in ACEs, if we require that it matched DAV:principal-URL on the principal resources, you can get that info in one request instead of two. I also get the feeling that having servers always vend the "canonical" DAV:principal-URL URI for the principal rather than being allowed to vend any of the alternates will avoid problems down the road, for the same reason that it is necessary that there be a canonical URL in the first place. But that's arguably a best- practices thing, not a necessity. But it's not possible to rely on if specs don't require it... -wsv
Received on Friday, 23 May 2008 16:55:38 UTC