In this snippet ```@Composable fun Toolbar(content: @Composable () -> Unit) { ... } fun slotBui...

# compose

[dimsuz]

In this snippet

@Composable
fun Toolbar(content: @Composable () -> Unit) { ... }

fun slotBuilder(text: String): @Composable () -> Unit = { Text(text) }

@Composable
fun Screen() {
  Toolbar { Text("hello") } // A
  Toolbar(content = slotBuilder("hello")) // B
}

Is A and B equivalent allocation-wise? In each case a new lambda will be created on each recomposition or am I wrong here?

[Ankit Kumar]

both a and b are same
and yes new instance of lambda will be created on each recomposition

[Loney Chou]

Lambdas will be remembered implicitly iirc.

[dimsuz]

Hmm, I do not recall reading about implicitly remembered lambdas anywhere. I guess I'll just check 🙂

[Ankit Kumar]

no i don't think it's remembered implicitly
but can wrap with remember

remember {
    {}
}

[Ankit Kumar]

I mostly prefer method ref to viewModel
but for some rare case I use above one

[dimsuz]

Can't wrap it with remember, because it's a composable function

[dimsuz]

Also if I'm not mistaken, debugging indicates that variant B has new lambda instance on each recomposition, while variant A has the same lambda instance on each recomposition. I'm not seeing clearly why this is so.

[Loney Chou]

Lambdas will be remembered implicitly iirc.

I was talking about A case. For B case, I don't know either 😕 Maybe because it's as return type.