Re: [ws-rx] Editorial issues with the latest editor's draft

[Doug Davis <dug@us.ibm.com>]

Matt - thanks - I applied most of them
- see inline for
a couple of comments.
Thanks to Paul for the wsse URL - saved
me the trouble of looking for it.
-Doug

Matthew Lovett <MLOVETT@uk.ibm.com>

07/28/2006 10:18 AM

To	ws-rx@lists.oasis-open.org
cc	Doug Davis/Raleigh/IBM@IBMUS
Subject	[ws-rx] Editorial issues with the latest editor's draft

Hi all,

Since we seem to be getting to the end of this process, here are a bunch
of editorial nits that we should fix.

Table 1, which lists the namespaces referenced in the spec, should include
wsse (as used in section 6.1). If we do not want to nominate a specific
level of the spec then we should list the alternatives. c.f. the 
definition for prefix 'S'.

Figure 1. When printed in black and white, the "Scope of RM Protocol"
text 
and circle are hard to read. Perhaps use a darker colour, if we can 
without obscuring the rest of the diagram.

try it now

3.1 Sequence Creation.
Both AcksTo and Offer/Endpoint are of wsa:EndpointReferenceType. Therefore
they both have element and attribute extensibility, as defined by WS-A.
However the exemplar (and following text) illustrates them differently.
My 
suggestion to the editors is that we strike the ellipsis from AcksTo, and
delete lines 318 - 320. 

Line 349 seems to be in the wrong font. Ditto 373.
 Good eye

We should move lines 362 - 372 above 356.

The create sequence response exemplar and text should be updated in line
with the other EndpointReferences. My suggestion is to delete the 
ellipsis, and strike lines 445-447.

3.2 Closing a sequence
Line 479 - 481. We have a repetition of 'wherever possible'. Please strike
the second occurrence, and surrounding punctuation.

3.6 Sequence acknowledgement
Move the text about None (lines 739 - 743) up to line 722 (e.g. just after
AcknowledgementRange)

3.7 MakeConnection
Line 773. I'm not sure if this reads better with an 's' -- "which
may be 
used to uniquely identify anonymous endpoint*s*", or make it singular
"which may be used to uniquely identify *an* anonymous endpoint".
Added the 's'

Line 776: Typo {uudi} for {uuid}

Section 5.2.2
Line 1201: Typo [WS-SecureConverstaion]
Line 1279: Typo missing space in "key(e.g.,"
Line 1289: Typo replace "an" with "and". Also insert
"the" after the comma 
in "headers, *the* fault behaviour"
Compare <wsrm:UsesSequenceSTR> and <wsrm:UseSequenceSSL> should
it be Use 
or Uses?  Went
with "Uses"


Comments on the schema file:
Line 204: Within OfferType, IncompleteSequenceBehaviour should hav 
minOccurs = 0
Line 37: Within SequenceAcknowledgement, Nack should not have a minOccurs
attribute  I'm
assuming you meant "None" - fixed
Line 69: Within MakeConnection, the attribute wildcard should not include
the 'use' attribute
I'm assuming you mean MessagePendingType
Line 84. In Identifier element remove the second attribute wildcard - it
has already been incuded on line 81.
Line 223: In Expires element remove the second attribute wildcard - it
has 
already been incuded on line 220.

Line 98/99: Within MakeConnectionType, the intent is one or the other,
or 
both. The schema doesn't capture that, should we put in a more complex
content model to make it happen? (Is it even possble to express in schema?
I have a feeling that this is what xs:all does, but I'm not sure)
We discussed this (perhaps
off-line) and decided that instead of making the schema really
complex we'd just add
text to the spec to do it.  We can revisit this decision if people
want.

Insert definitions of UsesSequenceSTR and UseSequenceSSL elements


Comments on the wsdl file:
This should be pulling in the new WS-A WSDL namespace, and replace 
wsa:Action with wsaw:Action. That might mean a second update to the 
namespaces table at the start of the spec. Perhaps we are waiting for this
to become a reccomendation?
 Dunno about this one.  Does anyone know if the 
wsaw namespace is fixed
or can it still change?

Thanks

Matt