> Anybody doing this on a production database without a strictly controlled (and tested!) break-fix should be fired on the spot. I agree with almost everything

> Anybody doing this on a production database without a strictly controlled (and tested!) break-fix should be fired on the spot. I agree with almost everything you said except this line. Usually, people with production access are (or should be) sensible enough not to touch the prod db with anything untested that might mutate things. If they're not sensible enough for that, they shouldn't have had access in the first place, and instead of placing blame and firing people, I'd start thinking about /why/ that happened. Points to a more fundamental flaw in the Engineering. Firing won't fix anything. Also note that the statement in question talks about an old methodology. As for the article, good stuff. I enjoyed reading it.