[jimrandomh]

You argue that the database should be thought of as a separate unit, and that unit tests should minimize interactions outside their unit. But this argument proves too much; it would also argue that unit tests should avoid using the memory allocator.

[sanderjd]

Your parent said "to the extent practical". Limiting interaction with external units as fundamental as the memory allocator is far from practical. I actually think DHH would agree with the "to the extent practical", but would think limiting DB access in tests of fundamentally DB-dependent things like ActiveRecord models is impractical. I think the debate is fundamentally about that practicality, and neither side is obviously right.

[joevandyk]

I think it's impractical to test things that use ActiveRecord without testing the database.

ActiveRecord at its core is meant to generate and run SQL and give you object graphs back, and you need to test that the SQL it generates is correct and does the right thing.

[dragonwriter]

> I think it's impractical to test things that use ActiveRecord without testing the database.

I think one of the fundamental, if not really directly addressed, divides between the sides here is over the question of whether domain models "using" (being tightly coupled to) the persistence layer is sensible architecture even outside of consequences for testing.

[sanderjd]

I hope my comment that you replied to doesn't make it sound like I disagree with you. I don't. But I do think it is reasonable for other people to disagree with us on that point. But I also agree with dragonwriter that perhaps a more important question is the practicality of having fewer things that "use ActiveRecord", such that you can then test those things without testing the database. I think it's a pretty good idea, but I've had only limited success doing it in practice.