[sailfast]

Things not considered by this and other articles that I would urge readers to consider: 1) RFPs are often written poorly by non-technical people, with requirements that are not accurate at the time. These requirements then change a lot to reflect reality which results in a lot of wasted effort and redirection. (This is probably also the case in large Enterprise implementations)

2) Compliance with government regulations costs money. Lots of money. This results in a lot more overhead. It also results in a lot more time to get people up to speed, on site, and going. This is why government contractors keep winning bids - compliance costs are huge barriers to entry.

3) Systems you need to integrate with in government (especially legacy systems) can be a complete pain in the butt. It's more likely you're integrating with some FORTRAN green screen than a nice JSON API. This makes large scale systems integration hard.

That said...the app is still very broken and there is obviously a failure here. Failure to test properly (otherwise poorly written tests), failure to open to competitive bids judging from another comment in this thread, and many other issues.

There is a LOT to be done to improve IT acquisition in government, and many things should have gone right that went wrong for the money spent (Figures I saw were more like $138 Million in other publications), but readers should please consider the organizational barriers and difficulties that exist and then factor it into the cost. It doesn't take the sting away, but it does lessen it a bit.