bitcoin The Future isn't Blockchain. It's HASHGRAPH.

[Frederik]

I wanted to introduce you guys to a technology which may change the future of our planet much more than Bitcoin will. It may even make Bitcoin obsolete.

If you know how Blockchain works, this topic should interest you.

–––––––––––

So...

Does expending huge amounts of resources to solve math problems that nobody really cares about make the world a better place?

Does "7 transactions a second maximum" sound like future world currency technology?

Is it fair that approximately every 10 minutes one single computer likely part of a supercluster wins the lottery and decides which transactions get to be included in the next block?

Is waiting for 1 hour for confirmation of a payment that even then isn't 100 % guaranteed sound viable in today's age, let alone in 20 years from now?

I'd say no, no, no, and no. But is there an alternative?

Yes. It's called Hashgraph.

–––––––––––

Hashgraph is a Distributed Ledger Technology much like Blockchain, yet very different – and much better.

In a nutshell, in Hashgraph every node (the client, like your computer running the program) tells random other nodes the data it has (transactions, coordinates, or other data), plus information it randomly received from other nodes. It's called "Gossip protocol" for obvious reasons and is not a new invention. So my node A sends my transaction to another node B, then A and B send to C and D, next round four nodes send out my information (exponential, so it only takes 20 iterations for 1 million nodes to know my transaction). Also, the nodes tell each other what they heard other nodes telling other nodes together with the timestamps, so they "gossip about gossip." The benefit of that gossip about gossip is that every node has all the data every other node has. This also demonstrates the fairness in Hashgraph, because all nodes are equal in power and no single node decides which transactions get to be included.

To decide on the correct order of the transactions, a voting system is used. But instead of sending votes all around in the network of who has seen which data when and to "vote for it", we can just forget about that part because everyone already knows how everyone "would" vote if there was a an actual vote – hence it's called "virtual voting" and is, together with the Gossip protocol, what makes Hashgraph incredibly efficient.

It's extremely efficient and fast, plus fair and secure. In fact, it's much more efficient, faster, fairer and securer than Blockchain! It's a straight upgrade.

It was invented 5 years ago by Leemon Baired of http://www.swirlds.com/ and they have only been public about this technology since September 2017.

–––––––––––

You want to learn more? See their website (above) or check out the following resources:

Hidden Secrets Of Money Episode 8 featuring Bitcoin and Hashgraph (video)

Overview of Swirlds Hashgraph (article)

How Hashgraph Consensus Works (presentation)

Swirds' Leemon Baird Talks Hashgraph (lecture)

–––––––––––

It may sound like I'm here to advertise, but I am of course in no way affiliated to them. I just have recently stumbled across Hashgraph in The Hidden Secrets of Money and have been learning much about it.

I find Hashgraph fascinating as this seems to solve all the problems that Bitcoin currently has. Most notably the extreme energy expense (between 1/2000th and 1/1200th of world's electricity!) and limited scalability.

Maybe you guys find this interesting, too? Have you heard about it before? What are your thoughts?

Cheers!

Edited January 3, 2018 by Fred Black Fox
Explanation added

[Kikker]

11 hours ago, Fred Black Fox said:

It may sound like I'm here to advertise, but I am of course in no way affiliated to them. I just have recently stumbled across Hashgraph in The Hidden Secrets of Money and have been learning much about it.

I find Hashgraph fascinating as this seems to solve all the problems that Bitcoin currently has. Most notably the extreme energy expense (between 1/2000th and 1/1200th of world's electricity!) and limited scalability.

Maybe you guys find this interesting, too? Have you heard about it before? What are your thoughts?

Cheers!

You sound like an advertisement because you keep presenting its advantages without actually explaining how the algorithm works and just citing videos (of one hour!) and the overview paper isn't gonna cut it. In other words your confidence in the advantages of the algorithm isn't reflected in the knowledge you display. I mean, do you know why the bitcoin is limited to and average of 4* transactions per second? How is it avoided in Hashgraph? Basic questions left unanswered! Why don't you give a short summary of it's inner workings if you already did the research?

Besides that, it looks interesting, it seemingly has it's roots in computational science instead of cryptology like block-chain. Though in the video it is very casually mentioned that the network can be disrupted when the hackers hold 33.4% of the nodes (though control is 66.8% of the nodes) while block-chain has a hard limit of 51% for both. Which sounds significant to me, especially since you don't have a constant push for computational proofs like in bitcoin meaning that inflating your control should be easier.

 

* a block is 2400 transactions and the algorithm controlling bitcoin keeps an average of 1 block every 10 minutes (2400/(10*60) = 4 per second)

[Frederik]

On 12/27/2017 at 11:17 PM, Kikker said:

You sound like an advertisement because you keep presenting its advantages without actually explaining how the algorithm works and just citing videos (of one hour!) and the overview paper isn't gonna cut it. In other words your confidence in the advantages of the algorithm isn't reflected in the knowledge you display. I mean, do you know why the bitcoin is limited to and average of 4* transactions per second? How is it avoided in Hashgraph? Basic questions left unanswered! Why don't you give a short summary of it's inner workings if you already did the research?

Besides that, it looks interesting, it seemingly has it's roots in computational science instead of cryptology like block-chain. Though in the video it is very casually mentioned that the network can be disrupted when the hackers hold 33.4% of the nodes (though control is 66.8% of the nodes) while block-chain has a hard limit of 51% for both. Which sounds significant to me, especially since you don't have a constant push for computational proofs like in bitcoin meaning that inflating your control should be easier.

 

* a block is 2400 transactions and the algorithm controlling bitcoin keeps an average of 1 block every 10 minutes (2400/(10*60) = 4 per second)

I guess it makes sense that my post would seem like an ad because I don't explain the technology.

In a nutshell, in Hashgraph every node (the client, like your computer running the program) tells random other nodes the data it has (transactions, coordinates, or other data), plus information it randomly received from other nodes. It's called "Gossip protocol" for obvious reasons and is not a new invention. So my node A sends my transaction to another node B, then A and B send to C and D, next round four nodes send out my information – exponentially, so it only takes 20 iterations for 1 million nodes to know my transaction which is the fastest possible way to distribute data in a network. Also, the nodes tell each other what they heard other nodes telling other nodes together with the timestamps, so they "gossip about gossip." The benefit of that gossip about gossip is that every node has all the data every other node has and that very quickly. This also demonstrates the fairness in Hashgraph, because all nodes are equal in power and no single node decides which transactions get to be included.

To decide on the correct order of the transactions, a voting system is used. But instead of sending votes all around in the network of who has seen which data when and to "vote for it", we can just forget about that part because everyone already knows how everyone "would" vote if there was a an actual vote – hence it's called "virtual voting" and is, together with the Gossip protocol, what makes Hashgraph incredibly efficient.

That's the basic gist of it I'll insert it into the opening post.

Now Hashgraph is much safer than Blockchain. While in blockchain there is never perfect certainty in the order of transactions, in Hashgraph there is 100 % certainty and that after a very short amount of time (seconds, not hours). Blockchain is not byzantine fault tolerant despite what people commonly believe. Hashgraph has the highest byzantine fault tolerance there is.

The 51 % number you quoted is actually false – every consensus algorithm can not guarantee integrity once more than one third of its participants have evil intentions. Leemon Baired explains why that is right here: CESC2017 - Leemon Baird - Hashgraph Security and Attack Resilience, 19m 54s

That should clear it up

[lorry]

It is more like statistics.

 

So they have a population of transactions, but it is held across a whole bunch of different nodes, and it evolves in time. So they just have the nodes at time t sample each other from time t - 1 and infer which node at time t - 1 has propagated the most information about the population at time t -1 into the population at time t by looking at how many times it was randomly sampled.

 

This is a weird system. Nodes don't have to do work like in bitcoin, they don't have to solve problems as such. The devs think they will save cpu work on solving math but in reality the work will then be done by running infinite virtual machines competing to be the node randomly sampled the most times. 

 

I don't like how they obfuscate what they are doing behind bullshit language, votes, famous nodes, seeing, strong seeing. I think they want to hide that getting your transaction approved is going to be completely random. I mean, if you can't just create multiple nodes with the same transactions on them so that your transaction gets sampled alot because it is already on alot of nodes so it will get approved faster (which, if it can be done, will be exactly what happens within 1 day) then your looking at uncertainty in the time to be approved that you can't pay a miner to fix (by including transaction fees in bitcoin).

[Tyler H]

The expenditure of electricity through mining is not a waste. The hashpower of the network secures it against malicious attackers. Any attacker would need to exceed the hashrate and power consumption of 51% of the network in order to manipulate it to their advantage. 

[lorry]

1 hour ago, Tyler H said:

The expenditure of electricity through mining is not a waste. The hashpower of the network secures it against malicious attackers. Any attacker would need to exceed the hashrate and power consumption of 51% of the network in order to manipulate it to their advantage. 

Probably. That is what I am thinking about hash graph. 

 

Bitcoin has 1 guy with 1 miner burning XYZ kWh hashing. Hashgraph has 1 guy with 1000 miners burning a total of XYZ kWh on random selection. What is the difference?

[Frederik]

22 hours ago, Tyler H said:

The expenditure of electricity through mining is not a waste. The hashpower of the network secures it against malicious attackers. Any attacker would need to exceed the hashrate and power consumption of 51% of the network in order to manipulate it to their advantage. 

Of course the work done in blockchain serves a purpose. That is, it serves a purpose in the consensus algorithm that is proof of work. That does not mean there is no possible other, better way to achieve the same level of security with other consensus algorithms.

The 51 % number is a myth, as I have pointed out above – it is 33 %. I see you were quick to post a reply

From my understanding, one reason for why it would be so hard to be malicious in Hashgraph is its immediate nature, in that data gets distributed immediately and therefore it would be very hard if not impossible to orchestrate a 33 % percent attack. That, and of course the nature of the Hashgraph, where every data block ("events") is cryptographically linked to all other events that is impossible to change later on just as it is impossible to change a block in Blockchain.

One big downside of proof of work is the problem of consolidation, in that nodes will move towards areas where it's cheaper to run a node (cost of electricity), and each will become bigger (economies of scale). When it's very inexpensive to run a full node however, one can expect that nodes will be running everywhere – from every ordinary PC to smartphones. That ensures that the system is backed up by enough trusted entities, plus makes it that much more fair.

Edited January 22, 2018 by Frederik
Elaboration added

[luxfelix]

I first heard about Hashgraph through the video posted by Mike Maloney as well.

Do you suppose that Blockchain and Hashgraph can coexist, each finding a niche it is better suited towards?

Or is Hashgraph -- hands down -- the next step in the evolution of distributed ledger technology?

[Frederik]

10 hours ago, luxfelix said:

I first heard about Hashgraph through the video posted by Mike Maloney as well.

Do you suppose that Blockchain and Hashgraph can coexist, each finding a niche it is better suited towards?

Or is Hashgraph -- hands down -- the next step in the evolution of distributed ledger technology?

I personally don't believe Blockchain will survive in the long term because it is very inefficient, slow and not Byzantine proof.

I kinda see Blockchain as the kickstarter of distributed ledger technology. In the same way that the first commercial motor vehicle wasn't all too efficient, I believe that the market will gravitate towards the more efficient solution.

And that might not be Hashgraph! But it will certainly be something very similar to it, because Hashgraph is so close to 100 % efficiency in terms of traffic needed to form consensus.

At the end of the day, with Blockchain you cannot have real currency since it is not practical to wait an hour or more for payment confirmation – Bitcoin is used more as an asset class than anything else. With Blockchain you cannot have games, because it is way too slow for that. I am not sure about auctions, because of the lack of fairness in Blockchain. Plus, the resource usage denies the possibility for Blockchain to become the basis for the Internet of Things, but that and so much more is what we will need in the future.