[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Re: [ws-rx] NEW ISSUE: What does 'anon' URI mean when used in AcksTo
Paul,
Taking a layered approach might make sense. I.e., Not define what 'anon'
AcksTo means in general and leave the interpretation to the
binding/transport of WSRM. But given that WSRM over SOAP/HTTP
implementation would be the most dominant case (at least in the
beginning), it would behoove us to define exactly what 'anon' AcksTo
means for SOAP/HTTP for the sake of interoperability (the example in
your email does indeed do that). I would still like to see what the
WS-Addr TC has to say about this.
-Anish
--
Paul Fremantle wrote:
> Anish
>
> There is a clear layering issue here. The SOAP core spec does not define
> the meaning of the anon URI because it is clearly meant to mean that the
> delivery of the message is up to the transport. So in the case of
> wsa:To, it means that by default of the transport, there is one and only
> one place to deliver that message - and for responses, that means using
> the HTTP response. In JMS, it most logically would mean the ReplyTo
> destination.
>
> Then - because SOAP has a binding, they define what that means when
> bound. We as a TC need to figure out if we want to delve into bindings,
> or remain at a higher level above SOAP only. So we could either define
> the binding, or we could specify it in terms of SOAP:
>
> For example, we could say that if the acksTo URI is anonymous, then the
> ack header should only be attached to a reply, and the reply should be
> in response to a message with a anon wsa:ReplyTo. And then it is up to
> the SOAP binding to define the binding.
>
> Paul
>
> Anish Karmarkar wrote:
>
>> As to the question of why one would use anonymous IRI for wsa:To:
>> WS-Addressing spec actually requires you to do so in certain cases.
>> Specifically, in the case of the req-res scenario where the
>> wsa:ReplyTo of the request message is an 'anon' IRI the value of
>> wsa:To of the response message must be 'anon' IRI. See [1]. 'anon' IRI
>> is also to default for wsa:To, so if there is no value specified then
>> it is 'anon' IRI.
>>
>> The use of 'anon' IRI in wsa:To message makes sense on the response
>> message in case of SOAP/HTTP. I don't know what it would mean (or why
>> it would be used) in other cases.
>>
>> Doug: I think this (what does 'anon' URI mean for wsa:To) is a good
>> catch from the perspective of WS-Addressing. It would be fruitful to
>> provide this as a CR comment.
>>
>> But the wsa:To issue is a more general issue whereas AcksTo is a
>> specific issue for WSRM.
>>
>> In resolving LC20 [2] WS-Addressing made a specific decision to not
>> define exactly what 'anon' IRI means in the Core spec and left it to
>> the binding specs. SOAP binding does define it for ReplyTo and
>> FaultTo, but not for others (I think this is an oversight, but don't
>> recall the full discussion). IMHO, it would be useful to have
>> WS-Addressing SOAP binding define (or provide guidance on) what 'anon'
>> IRI means for SOAP/HTTP in all cases and not just ReplyTo and FaultTo.
>> The EndpontReferenceType is meant to be reused by other specs (WSRM
>> has AcksTo, WS-Eventing has wse:NotifyTo, WS-Notification has
>> wsnt:SubscriptionReference, wsnt:ProducerReference etc) and having the
>> meaning of 'anon' IRI (which is very useful for SOAP/HTTP) for all
>> EndpointReferenceTypes (in the context of SOAP/HTTP) would serve a
>> very useful purpose.
>>
>> But I haven't thought it through, perhaps there are reasons why
>> defining the 'anon' IRI for SOAP/HTTP binding in every case does not
>> make sense -- in which case, we should define in WSRM spec what we
>> need (for AcksTo EPR).
>>
>> Comments?
>>
>> -Anish
>> --
>>
>> [1] http://www.w3.org/TR/2005/CR-ws-addr-core-20050817/#formreplymsg
>> [2] http://www.w3.org/2002/ws/addr/lc-issues/#lc20
>>
>> Doug Davis wrote:
>>
>>>
>>> Personally, I'm not sure what an anonymous wsa:To means either :-)
>>> but its allowed, and its actually the default value if wsa:To is not
>>> specified.
>>> The only similarity I was trying to make was that if WSA doesn't need
>>> to say what an anonymous wsa:To means then we might not need to say
>>> what an anonymous AcksTo
>>> means - it could just be obvious. But its 8pm on Friday and I might
>>> not be thinking straight since I'm also choosing to actually continue
>>> to work instead of doing something else....sigh
>>> thanks,
>>> -Doug
>>>
>>>
>>> "Yalcinalp, Umit" <umit.yalcinalp@sap.com> wrote on 09/09/2005
>>> 07:25:20 PM:
>>>
>>> > I fail to understand the similarity or more simply why one would
>>> use anonymous IRI
>>> > for "To". Where would you send the initial message? /dev/null ;-)
>>> > > AcksTo EPR using anon IRI is different, it is piggybacking on
>>> the existing channel
>>> > provided by the binding, similar to ReplyTo and Fault which
>>> corresponds to a kind
>>> > of "response". Acknowledgement is a kind of response.
>>> > > --umit
>>> > >
>>> > From: Doug Davis [mailto:dug@us.ibm.com]
>>> > Sent: Friday, Sep 09, 2005 3:47 PM
>>> > To: ws-rx@lists.oasis-open.org
>>> > Subject: RE: [ws-rx] NEW ISSUE: What does 'anon' URI mean when
>>> used in AcksTo
>>>
>>> >
>>> > Right - it talks about ReplyTo and FaultTo but not "To". So, we
>>> can either assume
>>> > that it is implicitly talking about "To" which means wecan make
>>> the same
>>> > assumption for AcksTo. Or we can assume that silence on "To"
>>> implies something
>>> > else - like "its obvious". dunno. Whatever that assumption is, we
>>> can probably
>>> > carry it over to the AcksTo EPR :-)
>>> > thanks,
>>> > -Doug
>>> >
>>> >
>>>
>>> >
>>> > "Yalcinalp, Umit" <umit.yalcinalp@sap.com>
>>> > 09/09/2005 06:24 PM
>>> >
>>> > To
>>> >
>>> > Doug Davis/Raleigh/IBM@IBMUS, <ws-rx@lists.oasis-open.org>
>>> >
>>> > cc
>>> >
>>> > Subject
>>> >
>>> > RE: [ws-rx] NEW ISSUE: What does 'anon' URI mean when used in AcksTo
>>> >
>>> >
>>> >
>>> >
>>> > Hmm. I am wondering whether we are looking at different versions
>>> of the soap
>>> > binding spec. The current editor's copy [1] has the following
>>> statement in Section
>>> > 3.5 (Anish was suggesting adding basically a similar language to
>>> our spec in this thread).
>>> > > {
>>> > When "http://www.w3.org/@@@@/@@/addressing/anonymous"; is specified
>>> as the address
>>> > of the ReplyTo or FaultTo EPR, the underlying SOAP protocol
>>> binding provides a
>>> > channel to the specified endpoint. Any underlying protocol binding
>>> supporting the
>>> > SOAP request-response message exchange pattern provides such a
>>> channel. For
>>> > instance, the SOAP 1.2 HTTP binding[SOAP 1.2 Part 2: Adjuncts]
>>> puts the reply
>>> > message in the HTTP response.
>>> > }
>>> > > > --umit
>>> > > [1]
>>> http://dev.w3.org/cvsweb/~checkout~/2004/ws/addressing/ws-addr-soap.html?
>>>
>>> > content-type=text/html;%20charset=utf-8#soaphttp
>>> >
>>> > From: Doug Davis [mailto:dug@us.ibm.com]
>>> > Sent: Friday, Sep 09, 2005 1:57 PM
>>> > To: ws-rx@lists.oasis-open.org
>>> > Subject: RE: [ws-rx] NEW ISSUE: What does 'anon' URI mean when
>>> used in AcksTo
>>> >
>>> >
>>> > I saw no mention of what an anonymous wsa:To means in the soap
>>> binding spec - but
>>> > perhaps I missed it. If not, then they're silent on it.
>>> > thanks,
>>> > -Doug
>>> >
>>>
>>> >
>>> > "Yalcinalp, Umit" <umit.yalcinalp@sap.com>
>>> > 09/09/2005 04:43 PM
>>> >
>>> > To
>>> >
>>> > Doug Davis/Raleigh/IBM@IBMUS, <ws-rx@lists.oasis-open.org>
>>> >
>>> > cc
>>> >
>>> > Subject
>>> >
>>> > RE: [ws-rx] NEW ISSUE: What does 'anon' URI mean when used in AcksTo
>>> >
>>> >
>>>
>>> >
>>> >
>>> >
>>> >
>>> >
>>> > Doug,
>>> > > I think it is incorrect to characterize that WS-Addressing is
>>> silent. It just
>>> > defers the definition to the binding where it belongs to the
>>> extent of how the
>>> > definition is used for WS-Addressing purposes only.
>>> > > I would be very much in favor of (a) or (b) for our own spec.
>>> If (b) can not be
>>> > coordinated with WS-Addressing wg (we need to do that rather fast
>>> due to the
>>> > timelines of WS-Addressing) , we should definitely explore (a).
>>> > > Cheers,
>>> > > --umit
>>> > >
>>> > From: Doug Davis [mailto:dug@us.ibm.com]
>>> > Sent: Friday, Sep 09, 2005 11:42 AM
>>> > To: ws-rx@lists.oasis-open.org
>>> > Subject: Re: [ws-rx] NEW ISSUE: What does 'anon' URI mean when
>>> used in AcksTo
>>> >
>>> >
>>> > Is WS-Addressing's silence on what it means to use the anonymous
>>> IRI in the wsa:To
>>> > header mean that it should be obvious to the reader? If so, then
>>> perhaps we can
>>> > take the same approach to its use in other places that
>>> WS-Addressing is silent as
>>> > well - for example, AcksTo - and say nothing and assume its
>>> obvious to the reader.
>>> > thanks
>>> > -Doug
>>> >
>>> >
>>> > Anish Karmarkar <Anish.Karmarkar@oracle.com> wrote on 09/09/2005
>>> 02:20:12 AM:
>>> >
>>> > > As discussed on the call today, I'm raising an issue about the
>>> meaning
>>> > > of 'anon' URI when used in AcksTo URI.
>>> > >
>>> > > Title:
>>> > >
>>> > > What does 'anon' URI mean when used in AcksTo EPR?
>>> > >
>>> > > Description:
>>> > >
>>> > > WS-Addressing Core [1], section 2.1 says the following about
>>> 'anon':
>>> > >
>>> > > "Some endpoints cannot be located with a meaningful IRI; this
>>> URI is
>>> > > used to allow such endpoints to send and receive messages. The
>>> precise
>>> > > meaning of this URI is defined by the binding of Addressing to a
>>> > > specific protocol."
>>> > >
>>> > > WS-Addressing SOAP binding [2] defines what the 'anon' address
>>> means
>>> > > when used with ReplyTo and FaultTo in SOAP and SOAP/HTTP
>>> binding. It
>>> > > does not say anything about what it means when used in other
>>> headers
>>> > > such as AcksTo.
>>> > >
>>> > > Justification:
>>> > >
>>> > > WSRM defines AcksTo element of type EndpointReferenceType and
>>> allows
>>> > > 'anon' URI for the address. But the meaning of such an anon
>>> address is
>>> > > not defined anywhere.
>>> > >
>>> > > Target:
>>> > >
>>> > > core, soap
>>> > >
>>> > > Type:
>>> > >
>>> > > design
>>> > >
>>> > > Proposal:
>>> > >
>>> > > This can be resolved by:
>>> > >
>>> > > a) Adding a stmt similar to WS-Addressing SOAP binding.
>>> Something like:
>>> > >
>>> > > "When "http://www.w3.org/2005/08/addressing/anonymous"; is
>>> specified as
>>> > > the address of the wsrm:AcksTo EPR, the underlying SOAP protocol
>>> binding
>>> > > provides a channel to the specified endpoint. Any underlying
>>> protocol
>>> > > binding supporting the SOAP request-response message exchange
>>> pattern
>>> > > provides such a channel. For instance, the SOAP 1.2 HTTP
>>> binding[SOAP
>>> > > 1.2 Part 2: Adjuncts] puts the reply message in the HTTP response."
>>> > >
>>> > > OR
>>> > >
>>> > > b) we could ask the WS-Addressing WG to fix their SOAP binding to
>>> > > include not just ReplyTo and FaultTo EPRs but any EPR when used
>>> in the
>>> > > context of SOAP/HTTP binding.
>>> > >
>>> > > I prefer that we do (b). If they refuse, we can do (a)
>>> > >
>>> > > Related issues:
>>> > > i012
>>> > >
>>> > >
>>
>>
>
>
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]