There are always a near infinite number of challenges with launching a new product; even more so if the product is conceptually different than the others in the space. Following the Theory of Constraints, you have to focus on the current constraining issue until it is resolved and you can move on to the next one.

With most new software products the constraints are tactical: code is late, functionality is limited, competition is too strong, execution is too difficult or expensive, and so on. None of those are true with DataRush. Our near term constraint is credibility. People don’t believe that we can do what we claim. Success stories are coming as we do more LightHouse Customer Projects, and endorsements are important as well, but technical people want to understand and believe.

So our best bet is to constantly remind architects and designers that:

• Dataflow is an accepted and proven approach that has a rigorous academic background, a history in chip design, and is the basis for other commercially successful efforts such as LabView from NI.
• We didn’t invent dataflow; all we are claiming is that we have written a Java implementation of a dataflow engine that takes advantage of these concepts.
• Just as there are a number of SQL implementations that have been developed over the years, we expect and even hope for additional dataflow implementations.

Here is one version of historical review of the major transitions in software over the last 30 years:

• from mini-computers to PCs,
• from command line to GUI,
• from single user to networked client-server apps,
• from multiple database approaches to SQL as a standard,
• from local networks to internetworked organizations,
• from occasionally connected to ubiquitous connectivity.

So just as hardware advances dictated software changes in the past, once again we see that the hardware transition from single core to multicore is dictating a switch from Von Neumann program counter code to inherently massively parallel architecture as seen in dataflow code...

UPDATE: We are so confident in the amazing speed and performance of our dataflow engine & library that we have decided to make a pretty aggressive offer, putting our money where our mouth is, so to speak. Check out the DataRush Challenge and take your 3 hour process to just 6 minutes!