Re: [ebxml-cppa] Constituent type (draft) modification for ebMS WSSconfiguration support (not relying on ws-policy based options)

["Monica J. Martin" <Monica.Martin@Sun.COM>]

Moberg Dale wrote:

>(quoted printable encoding trashed message original. Will send update
>before meeting.)
>
>Here is a modification to the current ConstitutentType to accommodate
>WSS  configuration support across multiparts, with both Mime part
>granularity and XML element granularity. This seems more than enough
>functionality to me, but add questions as appropriate. It still remains
>possible to use ws-policy for security to express configuration
>capabilities or agreements. Placed on agenda for discussion at August
>status meeting August 24 2007.
>
>
><xsd:complexType name=3D"ConstituentType">
>
>                        <xsd:sequence maxOccurs=3D"unbounded">
>
>                                    <xsd:element
>ref=3D"tns:SignatureTransforms" minOccurs=3D"0"/>
>
>                                    <xsd:element
>ref=3D"tns:EncryptionTransforms" minOccurs=3D"0"/>
>
>                                    <xsd:element ref=3D"tns:ElementRef"
>minOccurs=3D"0" maxOccurs=3D"unbounded"/>
>
>                        </xsd:sequence>
>
>                        <xsd:attribute name=3D"idref" type=3D"xsd:IDREF"
>use=3D"required"/>
>
>                        <xsd:attribute name=3D"excludedFromSignature"
>type=3D"xsd:boolean" default=3D"false"/>
>
>                        <xsd:attribute
>name=3D"excludedFromDataConfidentiality" type=3D"xsd:boolean" />
>
>                        <xsd:attribute name=3D"minOccurs"
>type=3D"xsd:nonNegativeInteger"/>
>
>                        <xsd:attribute name=3D"maxOccurs"
>type=3D"xsd:nonNegativeInteger"/>
>
>            </xsd:complexType>
>
>            <xsd:element name=3D"ElementRef"
>type=3D"tns:ElementRef"></xsd:element>
>
>            <xsd:complexType name=3D"ElementRef" >
>
>                        <xsd:attribute name=3D"signed" =
>type=3D"xsd:boolean"
>use=3D"optional"/>
>
>                        <xsd:attribute name=3D"encrypted"
>type=3D"xsd:boolean" use=3D"optional"/>
>
>                        <xsd:attribute name=3D"signBeforeEncrypt"
>type=3D"xsd:boolean" use=3D"optional" default=3D"true"/>
>
>            </xsd:complexType>
>
>=20
>
>=20
>
>Issues:
>
>=20
>
>1. Should the MIME part granularity attribute  "excludedFromSignature"
>have a default?
>  
>
mm1: Might be best as this is not optional use. We should think about 
whether default is practical for optional attributes however.

>2. Any combinations that need support left out?
>
mm1: If you also want to support security policy in some way, it also 
specifies encryptBeforeSign in Section 6.3 of specification. Thanks.