Agreed to add a statement that the inverse HasSubtype reference is mandatory and the forward reference is optional for DataTypes like it is already stated for other nodetypes.

Added statement to DataType NodeClass requiring inverse HasSubtype references

Agreed to text edited in telecon.

We need this also as errata for 1.04.
Clients need the inverse subtype information in a reliable way. We agreed that this was always the intention.

The final text in the 1.05 draft has some minor issues

• space at the end of the sentence
• "Servers need not provide HasSubtype References" sounds strange

Servers need not provide HasSubtype References, even if their DataTypes span a type hierarchy, however it is required that the subtype provides the inverse Reference to its supertype .

There is a related clarification needed for HasEncoding.

If a Client receives an ExtensionObject, the ExtensionObject.TypeId contains the Nodeid of the encoding object. To get the related DataType node, the client is doing an inverse browse with HasEncoding.

The specification is silent about the requirement of a inverse browse support. But this must be required too.

Added statement to HasEncoding description "The DataTypeEncoding Node shall provide the inverse HasEncoding Reference to its DataType."

Created Errata 1.04.10

Agreed to changes and 1.04 Errata in Virtual F2F.