- From: Tom Rutt <tom@coastin.com>
- Date: Tue, 08 Feb 2005 15:46:36 -0500
- To: paul.downey@bt.com
- CC: Savas.Parastatidis@newcastle.ac.uk, jmarsh@microsoft.com, public-ws-addressing@w3.org
paul.downey@bt.com wrote: >>Would you say the mapping from a logical wsa:to element to the way the >>message is transported is part of ws-addressing? >>If not, where is this mapping specified? >> >> > >somewhere else above and beyond this spec. > >i think there may indeed be systems which accept EPRs from unknown parties >and blindly send a reply, and that's fine. > > Well, what if the publisher of the epr wants to but the actual http url as the wsa:to address. How is this usage distinguished by the more abstract "logical" address use case you are championing. Perhaps we need a "choice" element to distinguish the two use cases. There will be many epr publishers which want to put a url to direct the http request to in the wsa:to field. Tom Rutt Fujitsu > >*but* many async scenarios involve some out of band negotialtion before the >messaging occurs. A sender may elect to only send messages to parties >known and use configuration to decide upon the actual means of transport. > >my company is keen to ensure that it's possible to build both styles of >architecture with this low-level component. > >Paul > > > > -- ---------------------------------------------------- Tom Rutt email: tom@coastin.com; trutt@us.fujitsu.com Tel: +1 732 801 5744 Fax: +1 732 774 5133
Received on Tuesday, 8 February 2005 20:47:33 UTC