I recently finished Scott Rosenberg's Dreaming in Code, which chronicles the ambition, start and middle of the Open Source Application Foundation's attempt to carry on the legacy of Lotus Agenda by creating Chandler, a world-changing personal information management system. The book is a terrific read and it was especially interesting to hear more about the fate of a system that I initially heard about many years ago at an Internet2 member meeting.
For me, the book highlighted the angst and turmoil that accompanies the search for greatness by design. I think most things that are truely world change are so do to the hapless ignorance of their creators, as setting out to make a contribution that is really of that scale brings an excess of hyperbolic baggage that makes it too scary or to difficult to accomplish much of anything. In the book, Rosenberg describes the very deliberate process of creating greatness, from a swanky SF office, to a series of long deliberations about the architecture of nearly every system component. In his account, it sounds as if the quest for doing something great displaced the ability of talented developers and hackers to simply make things. After all, "perfect" is the mortal enemy of "really good."
For anyone involved in the management of research or technical work, this book is highly recommended for its illustration of just how activity that looks like careful planning and management can hinder the emergent hacks that are critical for forward progress. Or, as Paul Graham described in an OSCon talk, "The atmosphere of the average office is to productivity what painting flames on the side of a car is to speed."
Via the local newspaper, I recently saw that Denmark once again ranks among the 