I’ve argued elsewhere that structural engineers are, in some ways, victims of our own success. in the mid-1800s, people were not shocked when a bridge or a building collapsed, particularly if it had been damaged by a natural disaster or by a fire. Today, people are generally surprised when a collapse of anything more substantial than a wood-frame single-family house occurs.
“Software is a Long Con” by Quinn Norton is not about structural engineering, but uses our success as a benchmark to attack the quality of commercial software. I’m not qualified to critique the argument, but I think it’s obvious that people accept software failures at a rate that would be considered catastrophic for structures. Part of the reason is the lack of serious consequences for most of the failures. If I lose ten minutes of work because the program I was using crashes, I get annoyed, and then I recreate the lost work, and then I move on. Low-level consequences like that, even if common, are simply less important than the life-threatening consequences from structural, civil, mechanical, or electrical engineering failures, to discuss just the world we live in at work.
In my opinion, the core of engineering as a profession is design, so I won’t say that software development isn’t engineering, but it feels like there aren’t enough professional safeguards against bad design.
You must be logged in to post a comment.