“Gossip about gossip” protocols [closed]

Lately there is much rumor about the (patented) hashgraph consensus algorithm, which claims to have very good complexity measures. See the whitepaper:

https://swirlds.com/downloads/SWIRLDS-TR-2016-01.pdf

A central part of this appoach, is the so called "gossip about gossip" which uses a gossip protocol to spred the gossip history itself through the network. At least, if I'm not mistaken. This builds a data structure called the hashgraph.

On top of this, then runs a Byzantine fault tollerance algorithm, but that is of no concern for this question, which is:

Is the hashgraph-algorithm really the first who uses such a data structure, to gossip about gossip, or are there previous/other approaches, too? Couldn't find anything, though.

Edit: I have problems finding the right Tags. Maybe someone could elaborate on this?

I think that all the voting algorithms are in a way doing gossip about gossip. It is not a revolution, you just share more about what you know and also how do you know it.

I suppose, Gossip about gossip is just a pitch. The real contribution is the math proofs on virtual voting based on the notion of seeing.

It is core of the patent - invention by Leemon Baird paper 16 Bird, patent