here's quote from synth:
The new consensus algorithm is beyond anything in any of the papers we have published. The new Skycoin consensus algorithm being implemented is significantly improved and different from what was published in the original white papers.

We are delaying publication of the new consensus algorithm and mechanism, because we do not want it stolen by other coins. We have to be ready with a large marketing push.

Otherwise, someone will just take it, rename it and do a shitcoin ICO on top of our work. We have already had coins in China try to clone the methods we have already published and do shitcoin ICOs on top of them.

The new consensus algorithm is amazing and bullet proof. It is so simple to implement, understand and simulate. It is just a complete revolution in distributed consensus.

We have perfected what Satoshi started.

We are even doing a pilot of running SQL databases on top of "blockchain" using this consensus algorithm, where each transaction is a commit modifying the data.

We are considering whether we want to file for a patent (with open source licensing of course), in order to stop copycats and shitcoins. In the US, if you invent something, someone else can patent it first and then sue you for using your own algorithm (first to file). So we have to consider our legal and marketing strategy for how we can release the consensus algorithm so that it is not copied.

I think towards the end of the year there will be a major release.

For instance, someone copyrighted the Skycoin logo and the logos of 60 other coins in one country and then can goto the coins and exchanges and shake them down for using the logo.

So we have to have the patents registered, the copyrights registered, etc to stop these patent and copyright trolls. Which is one reason we have not released more about it.

What Hash Graph has (and everything else we have evaluated), is just a poor man's version of what our algorithm can do with less complexity and faster. We have the only algorithm that has instant transactions with no 51% attack and which can be used unchanged for both public and private chains.

The algorithm is based upon mathematics, so the validity and convergence of the algorithm is not affected by the bugs in a particular implementation or the errors in a specific paper about a specific implementation of the algorithm.