Beautiful Teams

Capitulating

War correspondents in every conflict wire home stories of chauffeurs who try to avoid known roadblocks by taking detours through back country. Often the new route proves more deadly than the one they avoided. Such was the case with Longjump.

Speed on this project only caused time to accelerate, rather than to slow down as Einstein said it would. After only a month or two, our engineers realized that our new system components were not working together, and that software porting was going badly. We had no time to lose; we could hemorrhage at any roadblock. The investors insisted on meeting our schedule, so someone at the top laid down a radical shift of direction.

Instead of incorporating the RISC chip into a custom-designed and custom-built system, we would become a value-added retailer for another successful computer firm that was selling a computer based on that chip. Adopting their computer system would require only some scrambling around at the top of the device and software stack, rather than the intensive mixing and baking that the chefs of our engineering staff had originally planned.

Many of us felt our hearts sink upon hearing this strategy. Although it was certainly a blow to our pride to be using another vendor's product, we had substantive reasons for fearing the switch as well. Without our secret sauce, our craftsmanlike reengineering of the system from bottom to top, we would lose most of our performance advantages. We were, in effect, offering customers a ...