There are times when a system must adapt to new requirements within months or weeks rather than years. These new requirements can include complicated rules and new products or services the organization needs to scale to support. As the system scales up and becomes more complicated it can become very hard to adapt quickly to these changing requirements. In fact, the system can even become harder to change and slower to accept new requirements. This can lead to a desired architecture that is designed to scale allowing the system to more easily adapt to changing requirements.
Do you have the goal of building a system that can be extended and adapted without programmer intervention? Do you have the itch to explore meta-programming, not just because it is cool or complicated, but because you want your system's behavior or domain representations to scale? If so, consider learning about the design of systems that represent user-defined behavior specifications as metadata.
Architects create Adaptive Object-Model systems when they want to enable end-user programmers to adapt their system's behavior and they don't want developers to become the bottleneck. But what does it take to build a system that can be changed and adapted without programming? How do Adaptive Object-Models differ from little languages or DSLs and when is it appropriate to consider stepping into the meta world to build such an extensible system? This talk presents the basic ideas about meta-architecture such as an Adaptive Object-Model architecture and shares experiences that the presenters have had building and tuning various production meta-architecture systems that have enabled organizations to scale them more easily and without programming wizardry.
- Date
- March 30th, 2011