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

Re: Examples to Resolve Issue LC136

From: Tom Rutt <tom@coastin.com>
Date: Thu, 10 May 2007 20:20:58 -0400
To: David Hull <dmh@tibco.com>
Cc: WS-Addressing <public-ws-addressing@w3.org>
Message-id: <4643B6EA.6060607@coastin.com>

See inline

David Hull wrote:
> As far as I can tell:
> The first example works OK, except that as I understand it from the 
> WSP mailing list, alternative intersection is probably /bag union/, in 
> which case the intersected result is
> <ExactlyOne>
>   <All>
>     <Addressing>...</Addressing>
>     <Addressing>...</Addressing>
>     <jabber/>
>     <jabber/>
>   </All>
> </ExactlyOne>
> If it's bag /intersection/ (as the name suggests), then yes, you get 
> the results you gave, and intersection works here.  Even if it's bag 
> union, the result seems close enough.  We need to say what happens if 
> the same Addressing element (with the same children) appears more than 
> once anyway and I assume that "idempotence" or "set semantics" or 
> "ignore duplicates" or whatever we want to call it is the way to go"
> What's making it work is the definition of compatibility.  As long as 
> we assume that assertions are compatible iff they have the same 
> element name, then only the alternative with Addressing and jabber in 
> it will be compatible with itself.  The other pairs will be 
> incompatible and won't appear in the result.
> However, I don't think the second example works.  As I understand it, 
> you're using the alternative containing only Addressing(nonAnon) to 
> mean "or any (other) non-anon".  In that case, we're intersecting 
> "Non-anon http or non-anon jabber or any other non-anon" (which is 
> equivalent to "any non-anon") with "Non-anon mailto or non-anon jabber 
> or any other non-anon" (which is also equivalent to "any non-anon") 
> and getting "Non-anon jabber or any other non-anon" (which is again 
> equivalent to non-anon).
> That's not at all what I had in mind.  I had in mind one source (see 
> previous comment on "source" versus "client or server") saying "I know 
> that at least HTTP and jabber are allowed (but I don't know anything 
> else)", another saying "I know that at least mailto and jabber are 
> allowed (but I don't know anything else)" and the two together 
> therefore saying "we know that at least HTTP, mailto and jabber are 
> allowed (but we don't know anything else)".
The "i dont know anything else" is covered by the addressing with 
nonAnon nested assertion alternative, since that does not guaruantee 
that a uri
will work , it just says "some uri will work". 


Tom Rutt	email: tom@coastin.com; trutt@us.fujitsu.com
Tel: +1 732 801 5744          Fax: +1 732 774 5133
Received on Friday, 11 May 2007 00:51:26 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:04:16 UTC