LC305: notational conventions proposal

This issue proposes that we say in the pseudo-schema notational
conventions that extensibility points are not indicated, aligning with
WS-A which has that statement.

 

It appears to me that with such a statement, there would only be minor
abuse of the notational conventions, which is easily corrected.
Illustrating elements and attributes in another namespace (e.g.
xs:schema, wsoap:protocol) does not appear to be in conflict with the
general statement that extensibility points are not explicitly
indicated.  WS-A uses a convention of "xs:any" for a container which can
contain arbitrary element content, which seems better in-line with the
notational conventions than WSDL's "other extensibility elements"
annotation.  There is no convention to indicate mixed content (needed by
<wsdl:documentation>), so a local self-explanatory "extension" to the
notational conventions seems reasonable.  I also found a few errors that
need to be corrected.

 

Proposal:

1)     Reword WS-A Core section 1.1: "Pseudo schemas do not include
extensibility points for brevity." --> "For brevity, pseudo-schemas do
not indicate the existence of extensibility points."

2)     Likewise add the statement above to WSDL 2.0 Core section 1.4.8.

3)     Change pseudo-syntax in WSDL 2.0 Core section 3: "other extension
elements" -->"xs:any".

4)     Change pseudo-syntax in WSDL 2.0 Core section 5: "extension
elements" --> "mixed content".  This isn't strictly permitted by the
notational conventions, but is a self-explanatory.

5)     Consider adding pseudo-schema notation sample to WS-A.

6)     Add BNF-pseudo schema notation explanation, or a reference, to
Adjuncts.

7)     Correct typo in 5.6.3: wsoap:mepDefault="xs:anyURI ?" -->
wsoap:mepDefault="xs:anyURI"?.

8)     Correct typo in 5.8.3 and 5.1: wsoap:mep="xs:anyURI"? -->
wsoap:mep="xs:anyURI".

 

--  Jonathan Marsh  --  jmarsh@microsoft.com
<mailto:jmarsh@microsoft.com>   --
http://spaces.msn.com/members/auburnmarshes/
<http://spaces.msn.com/members/auburnmarshes/>   --

 

Received on Friday, 9 September 2005 17:46:22 UTC