- A namespaced object with an ownerReference referencing a uid of a namespaced kind which does not exist in the same namespace is now consistently treated as though that owner does not exist, and the child object is deleted.

For more information, see [#53040](https://github.com/kubernetes/kubernetes/issues/53040).

- Fixed incorrect handling by kubectl of custom resources whose Kind is "Status". ([#77368](https://github.com/kubernetes/kubernetes/pull/77368), [@liggitt](https://github.com/liggitt))
- Report cp errors consistently, providing full message whether copying to or from a pod.  ([#77010](https://github.com/kubernetes/kubernetes/pull/77010), [@soltysh](https://github.com/soltysh))
- Preserved existing namespace information in manifests when running `

- [`KT-64920`](https://youtrack.jetbrains.com/issue/KT-64920) Json.encodeToString yields different results depending on whether typealias is used
- [`KT-58260`](https://youtrack.jetbrains.com/issue/KT-58260) Make invoke convention work consistently with expected desugaring
- [`KT-67314`](https://youtrack.jetbrains.com/issue/KT-67314) PCLA works inconsistently with smart-cast related CS forks

- Apimachinery: Pretty printed JSON and YAML output is now indented consistently. ([#105466](https://github.com/kubernetes/kubernetes/pull/105466), [@liggitt](https://github.com/liggitt))

- Custom resource requests with `fieldValidation=Strict` consistently require `apiVersion` and `kind`, matching non-strict requests ([#109019](https://github.com/kubernetes/kubernetes/pull/109019), [@liggitt](https://github.com/liggitt))

    * The `kube-cloud-controller-manager` flag `--use-service-account-credentials` is now honored consistently, regardless of whether `--service-account-private-key-file` was specified.
* Fix credentials providers for docker sandbox image. ([#51870](https://github.com/kubernetes/kubernetes/pull/51870), [@feiskyer](https://github.com/feiskyer))

	// But for now any three digits.
	//
	// We used to send "HTTP/1.1 000 0" on the wire in responses but there's
	// no equivalent bogus thing we can realistically send in HTTP/2,
	// so we'll consistently panic instead and help people find their bugs
	// early. (We can't return an error from WriteHeader even if we wanted to.)
	if code < 100 || code > 999 {
		panic(fmt.Sprintf("invalid WriteHeader code %v", code))
	}