improve last README update about serializers and Kryo usage

README.md

@@ -37,8 +37,8 @@ import org.apache.flink.api.common.typeinfo.TypeInformation
 val eventStateDescriptor = new ValueStateDescriptor[Option[String]]("event",
   classOf[Option[String]])
 ```
-Although `flink-scala-api` does fallback to Kryo silently, above code example will eventually use Kryo. This may happen when old flink-scala-api is already disabled, but this Flink Scala API is not yet properly used: no serializers are used or imported from `org.apache.flinkx.api.serializers._`.
-Below code example is the right way to use Scala serializers coming from this library: it completely prevents Kryo from being used. Even if `implicitly` cannot find an appropriate TypeInformation instance for your type `T`, it will fail in compile time.
+Although `flink-scala-api` does not fallback to Kryo silently for Scala types, above code example will eventually use Kryo. This happens when official flink-scala-api (deprecated) is already disabled, but this Flink Scala API is not yet properly used: no serializers are used/imported from `org.apache.flinkx.api.serializers._`.
+Below code example is the right way to use Scala serializers coming from this Scala API: it completely prevents Kryo from being used. Even if `implicitly` cannot find an appropriate TypeInformation instance for your type `T`, it will fail in compile time.
 
 ```scala mdoc:reset-object
 import org.apache.flinkx.api.serializers._
@@ -49,8 +49,8 @@ import org.apache.flink.api.common.typeinfo.TypeInformation
 val eventStateDescriptor = new ValueStateDescriptor[Option[String]]("event",
   implicitly[TypeInformation[Option[String]]])
 ```
-
-See more on Java types serilization here: [java-types](#java-types)
+- See more on serialization with POJO serializer here: [using-a-pojo-only-flink-serialization-framework](#using-a-pojo-only-flink-serialization-framework)
+- See more on Java types serilization here: [java-types](#java-types)
 
 ## Usage