Re: [ebxml-msg] RE: pending issues

[Pete Wenzel <pete.wenzel@Sun.COM>]

>    > Section 5.1.3.8 on SOAP mustUnderstand reads too much like a
>    > tutorial.  Reword it by simply stating that this attribute is
>    > required on eb:Messaging header, and must have value "1".
> 
> <JD> The suggestion above is already implemented in 5.2 (eb:Messaging
> container element section in Packaging). So we can indeed remove the little
> "tutorial" section.

Ah yes...consider it gone.

> B. Issue #18 (SOAP Binding, Section 11-Old vs. Section 12-New).
>    Griping about standards processes (or lack thereof) and
>    rationalization of our direction in Section 12 (lines 3533-3554 in
>    WD 08) is unnecessary.  We need to specify, clearly and concisely,
>    exactly what we expect implementers to do, not why.  Propose to
>    strike these lines.
> 
> <JD> We Fujitsu agree - these llines were more for addressing previous
> comments . Maybe we could say something like: "Although a SOAP 1.2 version
> of SWA has not been formally submitted to W3C, it appears that most SOAP
> products have anticipated that usage, and after investigation, it appears
> that they have done so in a consistent, interoperable way. This
> specification is acknowledging these de-facto upgrades of SWA, which are
> summarized below."

Better.

>    The proposal to remove Section 11 would leave us with no normative
>    statements as to proper use of HTTP(S)/SMTP.  (Sections 12.2 & 12.3
>    contain only examples, not specifications of the bindings.)  If the
>    argument is that [SOAP 1.1 + WS-I BP (+ SwA)] or [SOAP 1.2 +
>    Adjuncts (+ SwA)], in addition to the HTTP, SMTP and related RFCs
>    contain all the detail necessary, then we should make such a
>    statement.  
> 
> <JD> According to the author (Hamid) who is also developer, the Sections in
> 12 have all necessary information for SMTP binding as well as HTTP. Except
> for the use of SOAPAction header that is required and could be reminded more
> explicitly (outside example). All useful statements are already added prior
> to the examples. All previous statements (section 11) are redundant with
> other referred specs (SWA, SOAP, RFCs.) and do not introduce any
> ebMS-specific profiling.

I will remove Section 11, and take it as homework either to confirm
this claim or raise specific issues.  If any other TC members have
concerns in this area, they should do the same.

-- 
Pete Wenzel <pete.wenzel@sun.com>
Sun Microsystems SOA & Business Integration
Standards & Product Strategy
+1 (626)471-6311, Sun x61311, US-Pacific TZ