In my last blog, I discussed the benefits of our abstraction layer. In this final blog in the series, I’ll wrap things up with a discussion of reusability and ease of use.

As you might imagine, our new paradigm for viewing data involves a number of new concepts that bear careful explanation. Issues such as different categories of the same metatype: inches and millimeters are really subclasses of length with a static relationship; Dollars and Euros are subclasses of currency with a dynamic relationship. As we have learned with object-oriented implementations, there are powerful benefits to hierarchical structures, but they can also render the system very complex and do not support our goal of creating an intuitive, easy to use system for a non-technical audience. We can certainly learn from object-oriented approaches how to balance between functionality and ease of use – a non-trivial endeavor but not an impossible one.

Another aspect of reusability and ease of use which our abstraction layer is architected to provide is the ability to create a relationship of a rule with a definition. The intent is for these rules to be applied to data in an automated fashion to align or cleanse it as the data is processed, without relying on the developer to repeatedly specify a set of rules. By removing these rules from the inline processing we can also optionally allow the rules to be managed independently of the data flow specification. In fact, we have demonstrated an approach similar to this with some indirect programming projects (sometimes referred to as meta-programming) completed for a number of our customers and prospects.

As we extend our abstraction layer, our goal is to provide an environment that is significantly easier to work in while at the same time providing a very rich set of functionality which can be used to process the data. We have introduced the concept of rules libraries in a recent release of the system, which allows complex algorithms to be abstracted in a way that they appear to be first-class functions within the environment (as if they were distributed as part of the system). This approach allows organizations to extend the capabilities of the system to provide increasingly complex functionality to less sophisticated users while retaining a consistent, easy to use interface.

Finally, we continually solicit feedback from our customers, partners, prospects, and industry analysts regarding both the data integration requirements of the business and how business users could become more closely involved in the delivery of solutions for their business problems. These discussions provide insights that drive our plans to introduce new functionality and ease of use. We welcome your comments.

expressor is excited about the possibilities for combining the ease of use of this semantic abstraction layer with the revolutionary cost of ownership model we have pioneered. Our approach is also attracting attention from the analyst community – Gartner Group recognized expressor with the “Cool Vendor” award for our innovative approach to data integration.

I hope you enjoyed this series – let us know what you think — and I look forward to exploring additional topics here in the future.

- Michael Ruland, field engineering