One of the main concern I've got during the implementation of PI AF is the representation / visualization of AF Element Attributes in the Clients and in the responses from the various Interfaces ( ODBC/JDBC, Restful web services .. etc ).
In many other applications we use , for example to manage product specifications, recipes, etc, the grouping and positioning of attributes/variables within an entity must respect certain rules, and the application provides a sort of "ordering field", that is automatically updated according the position of the attribute or attribute group / category when you physically create that attribute/variable within an element ( or element template ).
Now, in my specific use case, my element is an instrument ( or device as well ) that has attributes with three different categories that must be represented always in a specific order :
- Design parameters
- Engineering parameters
- Operating conditions / variables
Within this categories the attributes and sub-attributes must always be represented in the same order, ( as you enter it in the template ), and not by name.
In other applications I own and manage this kind of discussion started more than two decades ago, then it was implemented so far, and today it has been promoted as a best practice for the most common scientific and business application in the market.
I would like to know how some of you solved this issue without doing a very old and bad practice : putting numbers or alphanumeric characters in front of an element name, category name or attribute name.
Why we can simply have an ordering number field/property at least for category and attributes, that can be used for automatic sorting ( i.e. in sortField for PI webapi or PI SDK ? )
Thanks in advance for your comments and contribution for this topic.