W3C home > Mailing lists > Public > public-ws-addressing@w3.org > March 2005

Re: Why is [destination] defined as an IRI?

From: Anish Karmarkar <Anish.Karmarkar@oracle.com>
Date: Mon, 14 Mar 2005 11:47:30 -0800
Message-ID: <4235EA52.80603@oracle.com>
To: Savas Parastatidis <Savas.Parastatidis@newcastle.ac.uk>
CC: "Rogers, Tony" <Tony.Rogers@ca.com>, "Andreas Bjärlestam \\(HF/EAB\\)" <andreas.bjarlestam@ericsson.com>, public-ws-addressing@w3.org

That does not mean that properties (other than [destination]) are not 
required by the receiving endpoint. Reference Parameters are indeed 
required by the receiving endpoint, but they are not bound in 
[destination] (as it is an IRI); instead they are bound as separate SOAP 
header blocks (for the SOAP binding).
This means that any extensibility element defined in an EPR, if needed 
to be present in the message, has to define how it is bound. If 
[destination] were an EPR, then this may not be needed depending on the 
extensibility point. In fact, I happen to think that having 
[destination] an EPR and including the Reference Parameters in it is a 
better solution than binding Reference Parameters as separate SOAP 
header blocks because of problems with security and composability.

-Anish
--

Savas Parastatidis wrote:
> Tony,
> 
>  
> 
> You are correct.
> 
>  
> 
> An EPR contains a lot more information than just the address of the 
> receiver, which is what the [destination] is. An EPR may contain 
> metadata, reference parameters, etc. When you see a [reply endpoint] or 
> a [fault endpoint] you use the additional information in order to decide 
> how to interact with the described endpoint. Furthermore, an EPR tells 
> you, in the case of reference parameters, which headers you must include 
> in your message if you are to interact with that endpoint. There is no 
> need to include the metadata in your message; the endpoint already knows 
> the metadata about itself.
> 
>  
> 
> The value of the [destination] is the value of an EPR’s [Address].
> 
>  
> 
> Regards,
> 
> --
> Savas Parastatidis
> http://savas.parastatidis.name
>  
> 
> ------------------------------------------------------------------------
> 
> *From:* public-ws-addressing-request@w3.org 
> [mailto:public-ws-addressing-request@w3.org] *On Behalf Of *Rogers, Tony
> *Sent:* Thursday, March 10, 2005 6:27 AM
> *To:* Andreas Bjärlestam (HF/EAB); public-ws-addressing@w3.org
> *Subject:* RE: Why is [destination] defined as an IRI?
> 
>  
> 
> Just testing my own understanding, but I think this is because [source 
> endpoint], [reply endpoint], and [fault endpoint] each contain a 
> destination IRI of their own.
> 
>  
> 
> The destination is not the same kind of thing as the other three.
> 
>  
> 
> Putting it another way, the [destination]'s EPR is the one containing 
> all of these pieces.
> 
>  
> 
> Please, someone correct me if I've misunderstood.
> 
>  
> 
> Tony Rogers
> 
>     -----Original Message-----
>     *From:* public-ws-addressing-request@w3.org on behalf of Andreas
>     Bjärlestam (HF/EAB)
>     *Sent:* Thu 10-Mar-05 17:01
>     *To:* public-ws-addressing@w3.org
>     *Cc:*
>     *Subject:* Why is [destination] defined as an IRI?
> 
>      
> 
>     Why is the [destination] defined as an IRI while the [source
>     endpoint], [reply endpoint] and [fault endpoint] are defined as
>     endpoint references? Why can they not be treated equally? Is there a
>     fundamental difference?
> 
>     .Andreas
> 
Received on Monday, 14 March 2005 19:48:55 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 2 June 2009 18:35:04 GMT