W3C home > Mailing lists > Public > w3c-dist-auth@w3.org > January to March 2003

RE: Issues PUT_AND_INTERMEDIATE_COLLECTIONS and INTEROP_DELETE_AND_MULTISTATUS

From: Julian Reschke <julian.reschke@gmx.de>
Date: Fri, 24 Jan 2003 23:03:43 +0100
To: <w3c-dist-auth@w3c.org>
Message-ID: <JIEGINCHMLABHJBIGKBCGEPBGEAA.julian.reschke@gmx.de>

Well,

if we're trying to achieve consistency, we should try do it consistently.
Returning a multistatus with a non-207 status in just one specific case
seems like a workaround, not a fix. BTW: DELETE is always "depth
infinity" -- there is no shallow DELETE.

As far as I can tell, this would affect all namespace operations, such as
COPY, MOVE, DELETE and possibly LOCK. In general, this would enable clients
to do the HTTP-sanctioned range checks (status >= 200 && status <= 299).

Julian

--
<green/>bytes GmbH -- http://www.greenbytes.de -- tel:+492512807760

> -----Original Message-----
> From: Lisa Dusseault [mailto:lisa@xythos.com]
> Sent: Friday, January 24, 2003 10:49 PM
> To: 'Julian Reschke'; 'John DeSoi'; w3c-dist-auth@w3c.org
> Subject: RE: Issues PUT_AND_INTERMEDIATE_COLLECTIONS and
> INTEROP_DELETE_AND_MULTISTATUS
>
>
>
> > That being said, from a consistency p.o.v. I agree. I'll
> > assume for a moment
> > that few WebDAV clients indeed *do* evaluate a 207 on DELETE
> > (and other
> > candidates such as LOCK/COPY...), and those probably would be easy to
> > upgrade. If this is true, all we'd need to do is
> >
> > - deprecate status of 207 for failures
> > - introduce a new 4xx code such as INCOMPLETE OPERATION which
> > would carry
> > the same multistatus body
> >
> It doesn't seem nearly that big a change is necessary. The biggest
> change I'd consider given what I currently know, is that the spec would
> require that HTTP 1.1 method requests to non-collections should not
> result in 207. This would only affect the definition for DELETE, since
> that's the only HTTP 1.1 method that is defined defined as returning
> 207. DELETE could still return 207 in response to a Depth: infinity
> request to a collection.
>
> My reasoning is that HTTP clients address individual pages with DELETE
> requests, not collections.
>
> lisa
>
>
Received on Friday, 24 January 2003 17:04:30 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 2 June 2009 18:44:02 GMT