The Range data type uses a Double to define the lowest and a Double to define the highest value for Variables having Numeric data types. That works fine for most Numeric data types (Int32, UInt32, float, double) etc., but it does not allow to represent all values of Int64 and UInt64, because for values > 2^52 or <-(2^52) the double has not the same precision as an Int64 or UInt64.
Potential solutions:

• Clarifying text that for Int64 and UInt64 not all values can be specified and, in that case, the next lowest resp. next highest value shall be used.
• Define additional types and properties specific for Int64 and UInt64
• Other solutions?

The InstrumentRange Property is currently only defined as optional Property on AnalogItemType and ArrayItemType. Both have the mandatory EURange Property. To expose in OPC UA, that a variable has a limited range for its data type (e.g. Int6, UInt12, Int55) like used in many fieldbuses (e.g. IO-Link) without knowledge of the EURange it would be desirable to use the Property also on Variables having other VariableTypes. Therefore, it would be good to define this Property in general (in Part 3) like the EngineeringUnits Property.