4 ms·
This isn't his argument; he's arguing that people expect X to work for some set of cases (either because it's the latest hotness or because X's maintainers adve
by weberc2 10y ago
This isn't his argument; he's arguing that people expect X to work for some set of cases (either because it's the latest hotness or because X's maintainers advertise it for those cases) but it ends up being more work in those cases than the proven solution. This is different from "doesn't work well in all cases".
- carsongross 10y agoHe's arguing that ORMs break down for complex situations, so you shouldn't use them. It's a false dilemma. You can use an ORM for a lot of needs and then, when necessary, kick out to SQL. He says this "breaks the abstraction" but who cares? Abstractions aren't laws of physics: do what works. False dilemmas are the best dilemmas, because you can just ignore them. Unfortunately, programmers seem particularly susceptible to this sort of fallacy. I think it is due to a form of puritanism that I often see in the mathematical mind.