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

RE: Issues: MKCOL_AND_302, IMPLIED_LWS, PUT_AND_INTERMEDIATE_COLLECTIONS, INTEROP_DELETE_AND_MULTISTATUS

From: Julian Reschke <julian.reschke@gmx.de>
Date: Mon, 23 Jun 2003 13:30:11 +0200
To: "Lisa Dusseault" <lisa@xythos.com>, "Jason Crawford" <ccjason@us.ibm.com>, "'Webdav WG'" <w3c-dist-auth@w3c.org>
Message-ID: <JIEGINCHMLABHJBIGKBCKEFGHKAA.julian.reschke@gmx.de>

> From: w3c-dist-auth-request@w3.org
> [mailto:w3c-dist-auth-request@w3.org]On Behalf Of Lisa Dusseault
> Sent: Sunday, June 22, 2003 9:35 PM
> To: Jason Crawford; 'Webdav WG'
> Subject: Issues: MKCOL_AND_302, IMPLIED_LWS,
> PUT_AND_INTERMEDIATE_COLLECTIONS, INTEROP_DELETE_AND_MULTISTATUS
> 
> 
> 
> 
> 
> MKCOL_AND_302: This issue can be marked "inBis" if not CLOSED.  
> Draft -03 says that MKCOL can return 302 and there have been no 
> objections so far.

It doesn't say anything specific about MKCOL and redirects. Or am I missing something?

Speaking of which: section 12.1 talks only about 302 and 303 in Multistatus. It should talk about all 3xx codes, in particular 301.

> IMPLIED_LWS: This issue can be marked "inBis" if not CLOSED.  
> Draft -03 says that the HTTP rules are imported "including the 
> rules about implied linear white-space."

Agreed.

> PUT_AND_INTERMEDIATE_COLLECTIONS: This issue can be marked 
> "inBis" if not CLOSED.  Draft -03 says "The server MUST NOT 
> create those intermediate collections automatically.”

Agreed.
 
> INTEROP_DELETE_AND_MULTISTATUS: This is the old issue respecting 
> how HTTP clients might be confused by a 207 response to a DELETE 
> message, believing it to be a success message  
> <http://lists.w3.org/Archives/Public/w3c-dist-auth/1999AprJun/0062
> .html>.  Have we got consensus to continue using 207, on the 
> basis that by now it would break far more WebDAV clients to 
> *stop* using 207?
> 
> - Julian says continue using 207 but has also proposed switching to a 4XX
>   <http://lists.w3.org/Archives/Public/w3c-dist-auth/2003JanMar/0049.html>
>   <http://lists.w3.org/Archives/Public/w3c-dist-auth/2003JanMar/0065.html>
> - Roy argues it violates RFC2616  
>   
> <http://lists.w3.org/Archives/Public/w3c-dist-auth/2003JanMar/0046.html> 
> - My vote is to continue using 207
>   <http://lists.w3.org/Archives/Public/w3c-dist-auth/2003JanMar/0044.html>
> - The interim meeting attendees in Jan 2003 were unanimous in 
> continuing with 207
>   <http://lists.w3.org/Archives/Public/w3c-dist-auth/2003JanMar/0044.html>
> - John DeSoi points out that Netscape uses DELETE and 2XX should 
> not be redefined
>   <http://lists.w3.org/Archives/Public/w3c-dist-auth/2003JanMar/0057.html>
> - Bob Denny says let's not violate RFC2616
>   <http://lists.w3.org/Archives/Public/w3c-dist-auth/2003JanMar/0048.html>
> - Geoff Clemm might want to clarify his position
>   <http://lists.w3.org/Archives/Public/w3c-dist-auth/2003JanMar/0065.html>
> 
> I don't think we have consensus yet overall.  Please discuss, 
> clarify, or even simply restate your position.

I think Roy has made a very good point about clients that do DELETE and are not aware of 207 being a valid (error) code for DELETE. Thus my proposal is not to use 207 when the operation did not entirely succeed, but instead use a 4xx code (keeping the multistatus response body). I think at this point there are only few clients which will actually properly process a 207 on DELETE, and those can probably easily upgraded.

Julian


--
<green/>bytes GmbH -- http://www.greenbytes.de -- tel:+492512807760 
Received on Monday, 23 June 2003 07:30:24 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 2 June 2009 18:44:04 GMT