The article shows how to figure out where you are failing in a large government project. You would think that someone somewhere would learn that large monolith

The article shows how to figure out where you are failing in a large government project. You would think that someone somewhere would learn that large monolithic-fix-all-our-problems-and-replace-all-our-antiquated-systems-at-once projects are doomed to failure. There are enough examples out there to satisfy anyone. Having done modest by comparison government systems, I can tell you that you cant get everything down on paper at the beginning. You have to be prepared for change. Paperwork is the killer. Requirements are the killer. Design is the killer. They should start with something achievable and evolve into the desired system incrementally. The problem is that the tools necessary to achieve the task arent there. Government software is full of exceptions, special cases, union rules and all that stuff and we dont really have the tools to encapsulate the idea of special cases that can flow through the system.