The keywords "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC 2119 [].

When describing abstract data models, this specification uses the notational convention used by the XML Infoset []. Specifically, abstract property names always appear in square brackets (e.g., [some property]).

When describing concrete XML schemas [, ], this specification uses the notational convention of WS-Security []. Specifically, each member of an Element Information Item's [children] or [attributes] property is described using an XPath-like notation (e.g., /x:MyHeader/x:SomeProperty/@value1). The use of {any} indicates the presence of an element wildcard (<xs:any/>). The use of @{any} indicates the presence of an attribute wildcard (<xs:anyAttribute/>).

Where pseudo-schemas are provided for a component, they use BNF-style conventions for attributes and elements: "?" denotes optionality (i.e. zero or one occurrences), "*" denotes zero or more occurrences, "+" one or more occurrences, "[" and "]" are used to form groups, and "|" represents choice. Attributes are conventionally assigned a value which corresponds to their type, as defined in the normative schema. Elements with simple content are conventionally assigned a value which corresponds to the type of their content, as defined in the normative schema. Pseudo schemas do not include extensibility points for brevity.

When defining the cardinality of endpoint reference properties and message addressing properties, this specification uses the following notation: (n..m), where n is the minimum allowed number of occurrences of the property and m is the maximum allowed number of occurrences. When n has the same value as m then exactly that number of occurrences of the property must be present in the associated endpoint reference or message.

This specification uses a number of namespace prefixes throughout; they are listed in . Note that the choice of any namespace prefix is arbitrary and not semantically significant (see []).

WS-Addressing may be used with SOAP [, ] as described in &wsa-soap.title;[]. WS-Addressing may be used with WSDL [, ] described services as described in &wsa-wsdl.title;[]. Examples in this specification use an XML 1.0 [] representation but this is not a requirement.

All information items defined by this specification are identified by the XML namespace URI [] &nsuri;. A normative XML Schema [, ] document can be obtained by dereferencing the XML namespace URI.

An endpoint reference is a collection of abstract properties. This specification defines a core set of properties, but it is also possible for other specifications to extend these and/or add other properties. The semantics and XML Infoset representation for any such extension properties will be described in their defining specifications. An endpoint reference consists of the following abstract properties:

This section defines an XML Infoset-based representation for an endpoint reference as both an XML type (wsa:EndpointReferenceType) and as an XML element (<wsa:EndpointReference>). For brevity simple XML terms are used, e.g. 'element' instead of 'element information item' - this is not intended to constrain use of the constructs defined in this section to textual XML representations.

The wsa:EndpointReferenceType type is used wherever a Web service endpoint is referenced. The following describes the contents of this type:

The following describes the attributes and elements listed in the schema overview above:

The following shows an example endpoint reference. This element references the the endpoint at the URI "http://example.com/fabrikam/acct".

This specification provides no concept of endpoint identity and therefore does not provide any mechanism to determine equality or inequality of EPRs and does not specify the consequences of their equality or inequality. However, note that it is possible for other specifications to provide a comparison function that is applicable within a limited scope.

This specification does not define a lifecycle model for endpoint references and does not address the question of time-to-live for endpoint references. Other specifications that build on or use WS-Addressing may define a lifecycle model for endpoint references created according to that specification.

As noted in endpoint references are extensible. When extension attributes or elements appear as part of an endpoint reference, the processing model for such extensions is defined by the specification for those extensions. Software that processes endpoint references can safely ignore any such extensions that it does not recognize or understand.

Extension elements and attributes MAY add additional properties to an endpoint reference in addition to those specified in . Endpoint reference extensions MAY modify the value of one or more existing properties of an endpoint reference. Extensions MAY modify the rules for binding endpoint reference properties to message addressing properties, or otherwise indicate that a different binding be used.

Note that this ability to modify existing properties and binding behavior, when coupled with the fact that software can ignore unknown or unrecognized extensions, may result in a difference in behavior depending on whether such an extended endpoint reference is processed by software that understands the extension(s). When designing endpoint reference extensions designers should consider that standard processing per this specification will prevail in cases where their extension is not recognized or understood.

The Architecture of the World Wide Web, Volume One [] recommends [, Section 2] the use of URIs to identify resources. Using abstract properties of an EPR other than [destination] to identify resources is contrary to this recommendation. In certain circumstances, such a use of additional properties may be convenient or beneficial; however, when building systems, the benefits or convenience of identifying a resource using reference parameters should be carefully weighed against the benefits of identifying a resource solely by URI as explained in [, Section 2.1] of the Web Architecture.

Message addressing properties collectively augment a message with the following abstract properties to support one-way, request-response, and other interaction patterns:

The [destination] and [action] properties indicate the target processing location and the verb or intent of the message respectively. The values of these properties can be used to facilitate the dispatch of messages.

A binding of WS-Addressing message addressing properties MUST reflect the property cardinality shown above. &wsa-soap.title;[] defines such a binding for the SOAP [, ] protocol.

The following shows the XML Infoset representation of the message addressing properties defined in :

The following describes the attributes and elements listed in the schema overview above:

Note that each of the element information items described above allows attribute wildcards for future extensibility. A message processor may safely ignore any extension attribute it does not recognize. Attribute extensibility allows use of xml:id[] for identification of these elements if desired.

This section describes the process of constructing a message in accordance to an EPR.

This section specifies the WS-Addressing-specific rules for creating a reply or fault message related to another message.

The following example illustrates a message containing message addressing properties serialized as header blocks in a SOAP 1.2 message:

This message would have the following property values:

The following example illustrates a reply to the above message:

This message would have the following property values:

To prevent information disclosure, EPR issuers should not put sensitive information into the [address] or [reference parameters] properties unless it has been adequately protected against arbitrary disclosure.

Some processors may use [message id] as part of a uniqueness metric in order to detect message replay. Care should be taken to ensure that, for purposes of replay detection, [message id] is composed from data, such as a timestamp, such that a legitimate retransmission of the message is not confused with a replay attack. It is also advisable to use a [message id] that is not predictable, to prevent attackers from constructing and sending an unsolicited reply to a message without having to see the actual message.