# channelFlow in Kotlin 1.5

**URL:** <https://discuss.kotlinlang.org/t/channelflow-in-kotlin-1-5/22601>\
**Category:** Libraries\
**Created:** [August 5, 2021, 6:53am UTC](https://discuss.kotlinlang.org/t/channelflow-in-kotlin-1-5/22601 "2021-08-05T06:53:52Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![SdprWeb](https://avatars.discourse-cdn.com/v4/letter/s/48db29/32.png) [@SdprWeb](https://discuss.kotlinlang.org/u/SdprWeb)\
**Post date:** [August 5, 2021, 6:53am UTC](https://discuss.kotlinlang.org/t/channelflow-in-kotlin-1-5/22601/1 "2021-08-05T06:53:52Z")

</div>

Hi, I’m looking for clarification on the stability of the `channelFlow` coroutines API. In Kotlin 1.5.0, the `channelFlow` API was made stable:

```auto
public fun <T> channelFlow(@BuilderInference block: suspend ProducerScope<T>.() -> Unit): Flow<T> =
    ChannelFlowBuilder(block)

```

However, the `ProducerScope` API used as a lambda parameter argument remains experimental.

My question is; doesn’t this severely limit the utility of `channelFlow` if it’s not possible to use the `send` function supplied by the experimental `ProducerScope` without using experimental features? Am I missing something? Any answers appreciated!