- From: Doug Davis <dug@us.ibm.com>
- Date: Thu, 9 Feb 2006 16:12:18 -0500
- To: Mark Nottingham <mnot@mnot.net>
- Cc: Anish Karmarkar <Anish.Karmarkar@oracle.com>, Mark Little <mark.little@jboss.com>, public-ws-addressing@w3.org, public-ws-addressing-request@w3.org, Paul Fremantle <pzfreo@gmail.com>
- Message-ID: <OFDB6C159E.FC9FB033-ON85257110.007437D3-85257110.00744AB0@us.ibm.com>
Ah, but other WS-* specs are placing wsa:ReplyTo in the message with its own semantics - not that different from wsa:From, IMO. -Doug Mark Nottingham <mnot@mnot.net> Sent by: public-ws-addressing-request@w3.org 02/09/2006 04:08 PM To Anish Karmarkar <Anish.Karmarkar@oracle.com> cc Mark Little <mark.little@jboss.com>, Paul Fremantle <pzfreo@gmail.com>, public-ws-addressing@w3.org Subject Re: WSA From ReplyTo, FaultTo and To have more concrete semantics; they actually have use cases baked into the WS-Addressing spec family (i.e., they all have to accept a certain type of message; the reply, a fault, and the original message, respectively). From just floats out there... On 2006/02/09, at 12:23 PM, Anish Karmarkar wrote: > Mark, > > Isn't that a general problem that exists, not just with wsa:From > but with wsa:ReplyTo, wsa:FaultTo, and wsa:To as well? > > -Anish > -- > > Mark Nottingham wrote: >> Speaking just as me... >> It sounds like there are a lot of potential use cases for From. >> What's less than clear is whether they're compatible; i.e., if WS- >> Foo says wsa:From should contain a URI that corresponds to the >> MAC address of your ethernet controller, WS-Bar says wsa:From >> should contain a urn:uuid for your service (as we're already >> seeing from our friends from the North, apparently), and WS-Baz >> says it should be your IP address, how do you use these >> specifications in a "composable" fashion? >> Of course, From could changed to allow more than one URI, but >> then how do you pick which one is the appropriate one? E.g., if I >> see three http:// URIs in there, which one is MY From? >> Smashing a bunch of different use cases into one vague semantic >> bucket isn't interoperable; it's asking for trouble. I see no >> reason why these different cases can't specify different headers >> to contain the information they need; yes, WS-Addressing is one >> boat that they could hop onto on the way to standards paradise, >> but there are others. >> Cheers, >> On 2006/02/08, at 4:11 AM, Mark Little wrote: >>> >>> +1 >>> >>> I don't see what it adds in removing it, but I can see what it >>> removes by removing it. >>> >>> Mark. >>> >>> >>> Paul Fremantle wrote: >>> >>>> Hi >>>> >>>> I'd like to write in support of wsa:From. >>>> >>>> 1) A lot of mediation (SLA provision, security checks, etc) is >>>> based on who/where the message came from. From is useful for that. >>>> 2) WSA makes WS-* much more "peer-to-peer". But knowing where a >>>> message comes from is a key part of that. >>>> >>>> For example we in Apache Synapse are allowing users to do >>>> custom routing based on wsa:From. >>>> >>>> Paul >>>> >>>> -- >>>> Paul Fremantle >>>> VP/Technology, WSO2 and OASIS WS-RX TC Co-chair >>>> >>>> http://bloglines.com/blog/paulfremantle >>>> paul@wso2.com <mailto:paul@wso2.com> >>>> >>>> "Oxygenating the Web Service Platform", www.wso2.com <http:// >>>> www.wso2.com> >>> >>> >>> >> -- >> Mark Nottingham http://www.mnot.net/ > -- Mark Nottingham http://www.mnot.net/
Received on Thursday, 9 February 2006 21:10:51 UTC