W3C home > Mailing lists > Public > public-ws-addressing@w3.org > March 2007

Separate constraints for replies and faults

From: David Hull <dmh@tibco.com>
Date: Mon, 05 Mar 2007 15:12:36 -0500
Message-ID: <45EC79B4.5010203@tibco.com>
To: "public-ws-addressing@w3.org" <public-ws-addressing@w3.org>

I believe I saw a similar question float by recently, but I'm not sure I
saw the answer:

Suppose I want to say "I can send replies either to a third party or
back on the back-channel, but I will only send faults back on the
back-channel".  How would I say this?

>From reading through WS-Policy again, and from Chris's comments, I think
the swimming-downstream way to say this is

<all>
  <exactlyOne>
     anon is OK for replies
     non-anon is OK for replies
  </exactlyOne>
  <exactlyOne>
    anon is OK for faults
  </exactlyOne>
</all>

The normalization rules then expand this combinatorially into what we'd
expect.

Is this about right?  If so, just how does that come out in
fully-spelled-out angle brackets?  Frankly, I
find WS-Policy very confusing, but that's more a WS-P issue.
Received on Monday, 5 March 2007 20:13:21 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 2 June 2009 18:35:16 GMT