I find it hard to empathize with this article, for a lot of reasons, but mostly because I'm not sure if the premise is true. It is indeed very easy to write so

I find it hard to empathize with this article, for a lot of reasons, but mostly because I'm not sure if the premise is true. It is indeed very easy to write software that will later fail, and not know that it will fail while you're writing it. It's even very possible to have software that is failing, very publicly, and not know why. Aside from levying blame here (which I think falls to management for having changed scope within weeks before launch), the thing that most people seem unwilling or unable to accept is that the website itself is a black box to a legacy government system. As with all black boxes, there's absolutely no way to know how it's expected to behave under load, especially when it seems that nobody ever got around to load testing it. I would be very surprised to hear if there were those 'in the know' who could have scuttled the launch here because there wasn't any testing. Now, for sure, there very well ought to have been people saying "We can't launch without testing!", but even if there were people in that capacity, that would have surely seemed like fear-mongering, and could likely have gotten them cast as evil Tea Party members, given the public pressure (and political pressure) to launch. Let's not forget that the alternative would have been to delay the effects of the law, which is a debate that shut down our government for a time. The outcome here, in my opinion, doesn't levy blame on the software industry, but at worst, on software industry within the government. That system is indeed very broken, and I think blame appropriated upon it is warranted -- even though systems do succeed in spite of it.