Why use XContent instead of Jackson?

I see that XContent is an abstraction for handling content (cbor, yaml, smile, json), but Jackson already supports all these data formats.



Potentially you could add xml, protobuf, avro, thrift into the mix as well.
https://github.com/FasterXML?utf8=✓&query=jackson-dataformat

It would seem much cleaner if you just offered the POJO objects that need to be ser/de and let Jackson do it's magic, especially for the Java API where I find myself implementing POJO's representing the index mapping.