> Over and over again Acid RDBMS have proven that trying to "relax" the rules in pursuit of performance or worse, mystical holy grails that have never been right or correct for a primary datastore, is a mistake. And then people goes back to them, because is the best tool for ALL the primary data store jobs. ALL OF THEM.

On the contrary. The most successful RDBMS by far was MySQL in an era where it didn't have any kind of ACID (you could write the transaction keywords but they didn't do anything). As the story we're talking about now shows, RDBMSes are routinely deployed with transaction settings that their users don't understand, much less use; there are settings that would reduce bugs if anyone cared to use them, but no-one does. People cargo-cult the idea that they should be using an Acid RDBMS but they almost never actually want or need one.

The success there had more to do with it being a) free, at a time when there were far fewer of those options combined with that it b) ran on windows. The enterprise's pathological windows use for developer machines and the state of apples devices at the time and the near non-existence of linux enterprise laptops which continues today meant that they had the perfect product for that moment in history. I cannot overstate this: the people making these decisions think an acid transaction is a felony and not a term of art used by some of their employees.