Indeed. The single most expensive human error I ever made in my life, I made in a system that was architected in exactly this way that the author claims elimina

Indeed. The single most expensive human error I ever made in my life, I made in a system that was architected in exactly this way that the author claims eliminates the possibility of such errors. Yes, it is true that it was very easy to fix the problem and restore the system to working order. But, by the time we detected and resolved the defect, the system had already been operating out of spec for some time and the damage had already been done. This article seems to be ignorant of its own context. If ensuring that the database is internally consistent is really all it takes, then we all would have stopped worrying about it back in 1983 when Reuter and Harder originally formalized ACID.