• Greg talks about how we’ve failed our so completely that they now base their success on our always failing in the same way. He starts with your classic Hello World use-case, the common sex change
  • Greg talks about how we’ve forced our customers to work with data when they’re naturally behavior-centric
  • The problem with losing the historical record – we’ve lost the value of context and intent
  • Scott K asks about determining software behaviors by observing user behavior
  • Greg describes how Command Separation and the Event Sourcing pattern can help in solving this
  • K Scott asks about how this fits in with REST-ful architectures which are generally data-centric
  • Jon asks about the UI space efficiency of designing for behavioral interaction instead of data interaction
  • Some examples from HR: Jon likes to promote people, K Scott enjoys discussions of termination procedures
  • Kevin asks how what Greg’s proposing is different from task based UI’s we’ve already seen
  • Jon asks how to sell this to management, who sometimes doesn’t feel the need to share business process information with the software developers
  • Greg and K Scott talk about how data-centric style applications lose valuable context – educational tracking, shopping carts, medical records, and financial systems.
  • Scott K and Greg talk about how data-centric applications don’t handle histrory well. Greg points out that there’s a big difference between an event and a snapshot model.
  • Jon asks how we persist this kind of event information – do we need to move away from relational databases?
  • Greg talks about why the implementational details are less important than grasping the high level concepts.