Divergence from Scala 2 specification in case class desugaring

[julienrf]

I’ve noticed that the following code snippet does not compile in Scala 2, but does compile in Scala 3:

case class Foo(x: Int, s: String) {
  def copy(y: Boolean): Unit = ()
}

object Main extends App {
  val foo = Foo(42, "foo")
  println(foo.copy(x = foo.x + 1))
}

The reason is that in Scala 2, the compiler does not synthesize a method copy because there is already a method of that name that is defined by the user.

However, in Scala 3 the compiler does synthesize an overload of copy that works alongside the one defined by the user.

In the Scala 2 specification, it says that only the name is significant: https://scala-lang.org/files/archive/spec/2.13/05-classes-and-objects.html#case-classes

Is this difference from Scala 2 expected?