ebxml-msg — archive
[Date Prev]
| [Thread Prev]
| [Thread Next]
| [Date Next]
— [Date Index]
| [Thread Index]
| [Month Index]
| [List Home]
Re: non-ebXML links - T2
Marty,
I agree that the spec should indeed be more explicit. We should
probably also allow for other values to be used, prefereably
a URI that identifies a protocol other than ebXML RM and/or
transport-specific.
In the MS spec, we certainly cite HTTP and SMTP, but others
such as TCP/IP or UDP/IP (each with some additional "glue") at the
low end to MQSeries at the high end of the spectrum of possible
transport mechanisms are possible and allowed. We just haven't done explicit
bindings for them, nor is it likely that the TC will take this task
on for 1.1. Anyone is free to define a transport binding whether
for proprietary use or (hopefully) for non-proprietary use and
publish/register that binding in a manner befitting.
I see no reason whatsoever why an MQSeries "transport binding"
wouldn't be appropriate.
Cheers,
Chris
Martin W Sachs wrote:
>
> (I may be repeating myself but just in case...)
>
> OK. That makes sense but the spec. doesn't say it and should say it.
>
> One thing has to be clarified. There is a difference between "transport
> layer" and "external to the MSH". "External to the MSH" could be in a
> higher or lower layer. The IBM MQSeries example that people keep
> mentioning is a higher layer and would not usually be considered
> "transport". In the MS spec, "transport" is either SMTP or HTTP, neither
> of which are reliable.
>
> I would agree to the meaning of "external to the MSH", with suitable
> explanation, but would prefer that the value of the attribute be changed
> from "Transport" to something broader like "other", which would cover both
> HTTPR (when standardized) and reliability in a higher layer.
>
> Regards,
> Marty
>
> *************************************************************************************
>
> Martin W. Sachs
> IBM T. J. Watson Research Center
> P. O. B. 704
> Yorktown
Hts, NY 10598
> 914-784-7287; IBM tie line 863-7287
> Notes address: Martin W Sachs/Watson/IBM
> Internet address: mwsachs @ us.ibm.com
> *************************************************************************************
>
> christopher ferris <[email protected]>@Sun.COM on 09/17/2001
> 10:36:26 AM
>
> Sent by: [email protected]
>
>
To:
Martin W Sachs/Watson/IBM@IBMUS
> cc:
David Fischer <[email protected]>, Dan Weinreb
>
<[email protected]>, [email protected]
>
Subject: Re: non-ebXML links - T2
>
> All,
>
> The value "Transport" means that the reliable semantics of
> OnceAndOnlyOnce are provided for by the transport layer, not the
> MSH layer of software.
>
> It means that the enforcement of the deliverySemantics are
> provided for external to the MSH such that the ebXML RM
> protocol artifacts are not used/necessary.
>
> Cheers,
>
> Chris
>
> Martin W Sachs wrote:
> >
>
> The meaning of the value "Transport" is not defined for
>
> reliableMessagingMethod. I don't think there is universal agreement with
>
> your surmise. The usual meaning of the word "Transport" and the meaning
>
> intended almost everywhere in the MS spec. is the layer below ebXML-MS,
> and
>
> the only bindings that we have provided are for HTTP and SMTP, neither of
>
> which is reliable. If the team means that "Transport" means "Other"
> here,
>
> the spec should explain this point and, preferably, change the value to
> > "Other".
> >
>
> David Burdett take note. This is a comment on V1.1.
> >
>
> Regards,
>
> Marty
> >
> >
> *************************************************************************************
>
> >
>
> Martin W. Sachs
>
> IBM T. J. Watson Research Center
>
> P. O. B. 704
>
> Yorktown
Hts, NY 10598
>
> 914-784-7287; IBM tie line 863-7287
>
> Notes address: Martin W Sachs/Watson/IBM
>
> Internet address: mwsachs @ us.ibm.com
> >
> *************************************************************************************
>
> >
> > "David Fischer" <[email protected]> on 09/17/2001 12:55:30 AM
> >
>
>
To:
Martin W Sachs/Watson/IBM@IBMUS, "Dan Weinreb"
> <[email protected]>
>
> cc:
<[email protected]>
>
>
Subject: RE: non-ebXML links
> >
>
> My understanding is that a value of reliableMessagingMethod="Transport"
>
> essentially means Other. There are other RM methods which can produce
>
> onceandonlyonce and not be ebXML. This is not the same as
>
> deliverySemantics="BestEffort" which really means no RM.
> >
>
> David Fischer
>
> Drummond Group.
> >
> >
[Date Prev]
| [Thread Prev]
| [Thread Next]
| [Date Next]
— [Date Index]
| [Thread Index]
| [Month Index]
| [List Home]