Well, I had a three-day weekend right after the conference, and then a very busy week, but I’m finally getting around to posting some of the highlights of day three.

Air Bootcamp

In this 3.5-hour session, Mike Chambers and company walked us through the essential steps for building desktop applications using Flex Builder 3, Flash CS3, Dreamweaver CS3, and the command line compiler. I picked up a number of useful hints and tips, and came away even more impressed with the power and ease of use that Adobe has built into AIR (creating a desktop application from Flash CS3, for example, is insanely simple). However, there are still a couple of weaknesses. For example, Flex Builder 3 still does not support keeping a Web app and its corresponding AIR app in a single project; you need to create a separate project for each application, which can create maintenance headaches. Also, AIR printing is simply Flash printing, so we’re stuck with the accompanying limitations for the time being.

One particularly good piece of news - which was also mentioned in one of the general sessions - is that AIR now ships with database support (SQLite).

Beyond the Basics of LiveCycle Data Services

Jeff Vroom, Adobe’s Principal Scientist, presented a truly brain-filling discussion of LiveCycle Data Services. One useful tidbit is that managed associations are preferred over hierarchical values whenever IDs are available for the referenced objects. I’ll try to write more on this later, because it really merits its own post.

Flex Roadmap

Ely Greenfield blew us all away again. In this discussion of how the Flex team are re-thinking the framework in order to accommodate future innovation, he presented an idea that is simply brilliant. By reducing each component to its essence, and then implementing supplemental features and behaviors as discrete elements (which can often be re-used across multiple components), the framework can become virtually limitless in its potential for extensibility, customization, and innovation. The essence of a button, for example, is a label and a click event. Everything else (e.g. skin) can be left open to modification by the application developer. Using this simple premise, Ely demonstrated how the Flex framework might be changed to enable the application developer (or tool such as Thermo) to create a tremendously engaging UI control from a simple List component. This idea is inspirational not so much because of what it says about the future of Flex, but rather because of the ways we can apply this idea in all of our development work. For more thoughts along these lines, see Anatole Tartakovsky’s post.