Damn! No, that's right. I did a cvs checkout the
other day of an old version of CAF, looking for CAS stuff. I forgot to prune the
cvs source and check out from scratch!
Mark.
----- Original Message -----
Sent: Thursday, March 25, 2004 1:27
PM
Subject: Re: [ws-caf-editors] ALS
issue
What? According the schema the enlist-als element looks
like:
<xs:element name="enlist-als">
<xs:complexType>
<xs:complexContent>
<xs:extension
base="tns:AssertionWithProtocolURIType">
<xs:sequence>
<xs:element
name="als" type="xs:anyURI"/>
</xs:sequence>
</xs:extension>
</xs:complexContent>
</xs:complexType>
</xs:element>
Note that this is derived from an
AssertionWithProtocolURIType:
<xs:complexType
name="AssertionWithProtocolURIType">
<xs:complexContent>
<xs:extension
base="tns:AssertionType">
<xs:sequence>
<xs:element
name="protocol-uri" type="xs:anyURI"/>
</xs:sequence>
</xs:extension>
</xs:complexContent>
</xs:complexType>
We have assumed that the protocol URI
represents the protocol type. Are you saying this is wrong? I discussions with
other TC members, this appears to be a shared understanding.
The only
question we have is how the ALS Registrar knows when the complete set of
registrants is enlisted for a protocol structured from multiple ALSs. Our
implementation deals with this right now by simple assuming a one-to-one
correspondence, but I understand this is a limitation we (I) have
introduced.
Greg
Mark Little wrote:
This may be part of the global
ALS+protocol-type discussion and apologies if this is what other have been
saying, but I've been working under my "mental model" of what's in the spec.
without double checking. There's no way for an ALS to register what
protocol-type is wants to be associated with.
Mark.
---- Mark Little, Chief Architect,
Transactions, Arjuna Technologies Ltd. www.arjuna.com
|