Is this a bug or am I missing something? ```fun &l...

# announcements

[Dominaezzz]

Is this a bug or am I missing something?

fun <T> List<T>.idk(): List<T> = TODO()
fun <T> List<T>?.idk(): List<T>? = TODO()

fun lol() {
	val real: List<String> = emptyList()
	val empty: List<String>? = emptyList()
	
	real.idk() // Resolves the first function.
	empty.idk() // Only safe (?.) or non-null asserted (!!.) calls are allowed on a nullable receiver of type List<String>?
}

(Sorry about the names, they take up too much time to come up with)

This only happens when the receiver is generic. The functions resolve as expected for String .

[Milan Hruban]

The idk functions both have same JVM signature, so the second one is ignored I think. If you remove the last line ( empty.idk()....,) you should still get compilation error right?

[Marat Akhin]

Which Kotlin version are you using?

[Dominaezzz]

1.3.61

[Dominaezzz]

@Milan Hruban If the receiver is String , it compiles fine.

[Milan Hruban]

oh, it's because you are not on JVM, right?

[Dominaezzz]

Yes, sort of. It's common code.

[Marat Akhin]

Just as an experiment, could you try the same example in your environment with -Xnew-inference? Seems like a possible bug in type inference, maybe it's fixed in the new version.

[Dominaezzz]

Aiit

[Dominaezzz]

Ah I just remembered. New inference is enabled in the IDE by default.

[Dominaezzz]

I'll still try it anyway.

[Dominaezzz]

Yup,

kotlinOptions {
    freeCompilerArgs = listOf("-Xnew-inference")
}

fixes it.

[Marat Akhin]

I guess the reason for the discrepancy is that IDEA is funky with its MPP support w.r.t. new inference, that's why it shows as invalid code in the IDE, but successfully compiles from the terminal

[Dominaezzz]

Hopefully it's fully enabled in 1.4 or smth.

[Marat Akhin]

As far as I know, that's the plan 😃

[Dominaezzz]

Haven't read the blog in a while. 😅