When an AVEVA Adapter creates a new PI Point, it always uses the point class "classic", even if none of the attributes that are specific to this point class are ever used. Instead, AVEVA Adapters should create PI Points using the point class "base" unless we know that the attributes of another point class like "classic" are needed.
"classic" PI Points have the same attributes as "base" PI Points, plus more, so it is not a big deal to later change a "base" PI Point to a "classic" one (i.e. no configuration is lost), but changing a "classic" PI Point to a "base" one destroys the configuration of the "classic"-specific attributes. That is, "base" makes a good starting point and "classic" does not.
Also, the point class "classic" is just not user-friendly. In particular, locations 1 to 5, which are generic numbers that PI Interfaces could use to store configuration data to the PI Point, but the names of these attributes do not make it obvious what these numbers mean because they are generic. You need to refer to the documentation of the PI Interface to be sure. Yes, there are some general conventions, such as location 1 being the ID of the PI Interface instance and location4 being the scan class, but this is not guaranteed and this knowledge comes mainly from experience with multiple PI Interfaces. The location attributes are not beginner-friendly.
In my opinion, the point class "classic" should be left in the past with the PI Interfaces. PI Connectors never had to use "classic" PI Points since they always used "base" ones. If AVEVA Adapters need to store more data in PI Points than the "base" class allows, then perhaps a new, better point class should be introduced.
