Linked #5

One thing I came to think of when I read Linked: Barabási writes that in graph theory (the forerunner to network science) growth was never considered: they always worked with graphs with a fixed number of nodes (and perhaps even a fixed number of links, I can’t remember). As network science took off, they soon found that without growth you didn’t get the scale-free topologies observed in most networks, so growth is central to network thinking.

This made me think of how up-front design is about figuring out what the design should be, and that implementing this design is not about growth: the assumption isn’t that the system should grow to become the reality envisioned by the plans – instead the system is built according to the design documents. I think this is an important distinction: when you assume that growth will take place, the plans turn into visions of what the system might become; at the same time you assume that various things (requirements changes, market changes, new learning) will affect the final architecture.