Skepticism of Competence

My wife said something to me yesterday that I've been really thinking about now for the last 24 hours. She's worked in banking, logistics, teaching, and is now in tech. I brought up how software teams work on shared code using Git repositories, and each change has to be approved by other developers before merging. Even before merging, there is usually a QA team that actively tries to break the code post-change, and usually you must pass all the QA requirements before the merge is allowed.

This is in contrast, she said, to any other area she's worked. If you think about a team of accountants, they probably have a folder full of spreadsheets that they share. If an accountant changes something on one of the sheets, either a single value or adds a row or even adds a column, he/she usually can just do it. If a mistake is made, everyone has to huddle together to figure it out, but reverting the file is not usually possible without serious work.

The same is true, I'd have to think, for teams of doctors, teams of scientists, and so on. There is a level of trust between the team members that the people doing the work are competent. In the software world, the way things are setup is exactly the opposite: there is full skepticism that anything any developer does will work.

You can take this skepticism all the way to the top. I'd imagine Linus Torvalds hasn't just saved to the main branch of anything on his own in years. Do you think a lead accountant couldn't go into his/her team's files and change whatever? Of course!

Just an observation she made that kinda opened my eyes. We really do have systems and procedures in place to make sure people don't screw up. Who's right? I dunno.