Re: another approach to status codes, etc. in HTTP binding

Chris, Mark,

7/19/2001 6:02:43 PM, christopher ferris <chris.ferris@east.sun.com> wrote:

>Yeah, I was puzzled by this as well. The Expect usage has the server
>responding with a 417 Expectation Failed status which is in the
>Client Error range of status codes, not the server range. Therefore, 
>applying that to the mU faulting analogy in SOAP, a 4xx status would 
>seem more suitable than a 500.
>
>So, Mark B, are you suggesting that the mU fault should be reported
>with a 4xx status or are you suggesting something completely different?

That's exactly what I am suggesting, as I described several weeks ago;

http://lists.w3.org/Archives/Public/xml-dist-app/2001Jun/0017.html

>My take is that it is the client that erred in expecting the server
>to support some feature, not the otherway round. 

Right-o.

>Mark Nottingham wrote:
>> 
>> How does Expect help in this situation?

Expect has basically identical semantics to mustUnderstand.  The only difference being that mustUnderstand is 
explicitly associated with a header block, whereas Expect isn't associated with anything in particular.

MB

Received on Thursday, 19 July 2001 21:17:59 UTC