- From: Mark Baker <distobj@acm.org>
- Date: Thu, 27 Jan 2005 10:44:17 -0500
- To: "public-ws-addressing@w3.org" <public-ws-addressing@w3.org>
On Thu, Jan 27, 2005 at 09:23:17AM -0500, Francisco Curbera wrote: > Hi Mark, > > You write, > > > What I understood Paco to want is to be able to *identify* services > behind a gateway using RefParams in the request message. > > You seem to be have joined the wrong battle. If you read carefully this > mail thread you'll realize I am just asking to get rid of the remains of > Section 2.3. I have been fighting since October to remove the work > "identifier" from the spec (and I am glad the group finally decided to do > so). I understand that there's some disagreement about what "identify" means, so I apologize for using that word. But earlier you wrote; ] The concern is that it is not an uncommon situation ] in real world enterprise deployments to use the same external address for ] more than a single service/configuration, so it is not a great idea to put ] those scenarios out of the scope. -- http://lists.w3.org/Archives/Public/public-ws-addressing/2005Jan/0148.html So let me ask; is it your intent that the requester be able to "target" a message to any particular service behind the gateway? If so, then my arguments about such an approach breaking layering apply. If not, then we're cool, though I don't understand why you'd then be concerned about "metadata", since the metadata should be that of the gateway itself, not of anything behind it (though the gateway is free to "aggregate" metadata of the services behind it if it wishes). Mark.
Received on Thursday, 27 January 2005 15:44:13 UTC