- From: David Hull <dmh@tibco.com>
- Date: Fri, 11 May 2007 10:32:48 -0400
- To: tom@coastin.com
- CC: WS-Addressing <public-ws-addressing@w3.org>
Tom Rutt wrote: > see inline > > David Hull wrote: >> Tom, >> >> Before I dig into this in detail, I wanted to point out one thing, which >> may or may not matter and is probably more a WS-P issue if it does. >> >> You say: "I changed the examples to a server policy intersecting with a >> client policy." >> >> I'm not sure that this is a safe change. >> > I do not understand your scenario. How can there there two policy > expressions for the same server that are not the same? I thought I gave three examples below? > > Tom >> I phrased the original question in terms of two sources of metadata. >> One knows something. Another knows something else. We want to combine >> that knowledge and figure out what they both know together. For >> example, an abstract WSDL says something general and a concrete WSDL >> provides more detail. Or we know some things because they're common to >> all servers that implement spec X and we know other things because >> they're company policy around here. Or there's a minimum service-level >> agreement (e.g., you must at least support anon, or you must not support >> anon and but you must at least support HTTP callbacks) and the server >> provides more capabilities (it supports anon and non anon, or it >> supports both HTTP and Jabber callbacks). >> >> This seems more general than "the client wants this, the server provides >> that", though if the WS-P intersection algorithm serves for both, great. >> >> Tom Rutt wrote: >> >>> This body part will be downloaded on demand. >>> >> >> >> >
Received on Friday, 11 May 2007 14:33:10 UTC