- From: Lisa Dusseault <lisa@osafoundation.org>
- Date: Fri, 28 Oct 2005 15:23:34 -0700
- To: Jim Whitehead <ejw@soe.ucsc.edu>
- Cc: WebDav WG <w3c-dist-auth@w3.org>
On Oct 28, 2005, at 3:13 PM, Jim Whitehead wrote: >> What if I remove the header from the response example, in addition to >> Jim's suggested change? >> > > Urk, bad idea. > > The use of Location with MOVE is to be consistent with the semantics > of the 201 response code. I read this a little differently than you -- I don't see that Location is REQUIRED with the 201 response code. So I don't see a problem with leaving it out of a 201 Created response to a MOVE, but I'm not adamant either way. > > So, the suggested text has to be more precise. Perhaps: > > "The use of the Location header with the 207 status code is > intentionally undefined." > > However, even this I'm starting to feel is a bad idea. Why do we stop > at Location? Why not create a table with all HTTP headers and state > whether they should be used with 207? > Indeed. It's certainly more work, but this kind of thing can be extremely valuable to implementors. As we can see from the discussion over the list of response codes that could and couldn't be used in 207 responses, it's a fine line to walk and quite possible to overspecify. Perhaps if we had a table where many of the cell values were simply "undefined" instead of "MUST NOT use" we'd have a better compromise... Lisa
Received on Friday, 28 October 2005 22:23:47 UTC