- From: Marc Hadley <Marc.Hadley@Sun.COM>
- Date: Wed, 24 Nov 2004 11:38:52 -0500
- To: Christopher B Ferris <chrisfer@us.ibm.com>
- Cc: Mark Baker <distobj@acm.org>, public-ws-addressing@w3.org
On Nov 24, 2004, at 11:01 AM, Christopher B Ferris wrote: > > Please see Paco's recent missive[1]... the EPR is NOT an identifier, > it is > an addressable reference. > The distinction isn't that clear to me. > The ref props/params *can* be used to provide additional information > that > the service provider will use > as it sees fit. One such purpose that has been used by WS-RF has been > to > pass keys/identifiers to There's that identifier word again. > resources (implied resource pattern) as ref props, but that is not the > only use of ref props/params. > In the context of the implied resource pattern, the ref props > serialized > as SOAP headers can be > considered the equivalent of cookies used to associate a stateful > session, > like a shopping cart service > might do. > That may be your view but it is not clearly reflected in the specification status quo. In particular, ref params seems closer to cookies and the spec states that [address] + [reference properties] == identifier: "[address] : URI (mandatory) - An address URI that identifies the endpoint. This may be a network address or a logical address. [reference properties] : xs:any (0..unbounded) - A reference may contain a number of individual properties that are required to identify the entity or resource being conveyed." Note the use of 'identify' in each of the above. > As an example that is often used, a service might have three levels of > service; silver, gold and platinum. > Each level of service might have a different policy that applies. > Hence, I > would use the ref props to > include a <myservice:MembershipLevel> element with the possible values > Silver, Gold, or Platinum. > Is that identity? Nope. > It could be viewed in two ways: (i) You have the Gold Service, the Silver Service and the Platinum Service. The [address] + [reference properties] identifies the service you are using. (ii) You have one service that is parameterized by service level. The [address] identifies the service you are using. Using the status quo of the spec you'd achieve (i) by specifying a precious metal in a reference property and (ii) by specifying it in a reference parameter. Marc. > [1] > http://lists.w3.org/Archives/Public/public-ws-addressing/2004Nov/ > 0355.html > > Cheers, > > Christopher Ferris > STSM, Emerging e-business Industry Architecture > email: chrisfer@us.ibm.com > blog: http://webpages.charter.net/chrisfer/blog.html > phone: +1 508 377 9295 > > public-ws-addressing-request@w3.org wrote on 11/24/2004 09:48:35 AM: > >> >> On Tue, Nov 23, 2004 at 11:28:50AM -0500, Christopher B Ferris wrote: >>> >>> Dims, >>> >>> Why? There is no utility in making such a distinction from the > perspective >>> of a received message, they are >>> simply SOAP header blocks that are processed in the usual manner >>> using > the >>> SOAP processing model. >> >> Has the WG decided what the identifier is yet? Because if it has, I >> maintain that it's maximally self-descriptive for the identifier to be >> able to be located within the message which provides increased >> visibility for (generally) very little cost. Some might recall the >> issue with HTTP 1.0 allowing partial URIs in the request line, and the >> ensuing problems for supporting virtual hosting. This necessitated >> the >> introduction of the Host header in HTTP 1.1 which restored the lost >> identifying information. >> >> If the URI + RefProps is the identifier, then the RefProps need to be >> declared as such. If it's just the URI, then all is good; wsa:To >> suffices. If it's the whole EPR, then you need a way to distinguish >> RefProps & RefParams from each other, as well as other SOAP headers. >> >> Self-description means never having to say you're sorry. >> >> Mark. >> -- >> Mark Baker. Ottawa, Ontario, CANADA. http://www.markbaker.ca >> > > > --- Marc Hadley <marc.hadley at sun.com> Web Technologies and Standards, Sun Microsystems.
Received on Wednesday, 24 November 2004 16:38:56 UTC