[jasonpriestley]

The author suggests that the enormous economic success of software suggests that the methods used are not "crap", but I don't think this is valid. In fact, it's somewhat the opposite - the strong economic growth in the software world (and the exponentially improving performance and reach of the underlying hardware) allow low-quality software to proliferate.

The economic argument for quality engineering is, "it may cost more upfront, but will deliver a more reliable product with lower total costs." But this argument is invalid in the current software world, where network effects and first-mover advantage are enormous, and a successful company can grow explosively. No one would follow the strategy of "build a quick bridge out of plywood, then when we have 10 million people using it and it starts falling down every other day, we'll have enough money to hire some people to build a real bridge." But that's what twitter did.

As software developers, a lot of our job is just putting up with the crap that the software world has been built on, because the people in charge have made the rational calculation that it's better to hire ten times more engineers and expand to Asia, expand to mobile, expand up market and down market and into different markets a little sooner.

I agree entirely though, that it's counterproductive for programmers to blame ourselves or to seek a technical fix to what is basically an economic and organizational problem.