So, Will Fusion Really be a Superset of Product Features?

In their published white paper "Oracle Applications: Committed to Your Success", Oracle states that "Because Fusion is a “superset of features,”with a product-feature genealogy from Oracle,PeopleSoft, and JD Edwards, a much richer set of industry-specific features is now available in the release." The High-Tech defintion of a superset is "A set that includes other sets within it, which are called subsets. For example, a software or hardware upgrade may be a superset of the previous version in that it can do everything the previous version can do and more." So, by definition, Fusion applications should do everything that existing PeopleSoft, JD Edwards, and Oracle eBusiness applications can do. But is this true, or even possible?

Let's take a look at a feature that exists within PeopleSoft applications that does not exist with Oracle eBusiness applications. I'm referring to SetID. The concept of SetID allows multiple Business Units to share a common set of setup values while others do not. So, for example, PeopleSoft Purchasing Business Units shares information with PeopleSoft Payables Business Units that have the same SetID. With two SetIDs, you could have two sets of Purchasing and Payables Business Units existing side-by-side and not conflicting with each other. Each set of Business Units could have a Vendor ID 10001, but with each of these assigned to a separate SetID, there would be no conflict. Similarly, each set of Business Units would be sharing a completely different set of ChartFields. Oracle eBusiness applications have nothing comparable to SetID. While I'm not an expert on eBusiness, I do not believe that there is any clean way to have the type of scenario that I've just outlined in a single eBusiness instance. So, how will Fusion handle this? If SetID doesn't exist within Fusion applications, then PeopleSoft Enterprise that rely on this feature will be left out in the cold. Oracle has stated that the eBusiness data model will be used as the foundation for Project Fusion. Adapting this data model to include SetID will be an enormous task, given how prevelent it is in the PeopleSoft data model. Is is reasonable to expect that SetID will be added throughout the data model in order to accomodate PeopleSoft Enterprise customers that need it?

But we don't need to speculate to see that Oracle is already planning cutting back on features. Here is one example that shows that Oracle is clearly not committed to including all PeopleSoft features in the Fusion applications. PeopleSoft Order Management provides users with the ability to key in a One-Time address when entering an order to ship product to a location that is different from any of the ordering customer's existing addresses. This is commonly done if the customer is acting as a reseller and the order is being drop shipped directly to the customer's customer. Oracle applications do not have this feature. Any shipping address must be permenantly tied to the Customer Master. Oracle has no way to ship an order to an address that does not exist in the customer database. Now, if Fusion is truly going to be a superset, then the Order Management system in Fusion must support the concept of a One-Time address. There is currently a flurry of discussion amongst users of PeopleSoft Order Management due to the fact that it has been suggested to them that the One-Time address feature may be excluded from Fusion. PeopleSoft users that rely on this feature are finding that they are being forced to defend their use of this feature and to explain why it would impact their business if they were to lose it. We know that there are many more companies using eBusiness Order Management than there are using PeopleSoft Order Management. What this sounds like to me is that, in this case, rather than planning a "superset" release, Oracle is planning a "majority rules" release.

The loss of features that are key to some customers is probably inevitable if Oracle is going to keep to its announced schedule. It isn't realistic to expect that a true superset can be developed in the two to three year timeline that Oracle has given itself to bring out the first generation of Fusion applications. However, if Oracle doesn't own up to this fact with its customers, many of them are probably in for a rude awakening when the first release of Fusion becomes available.