Hello Michael. I believe your analysis is flawed because you missed one important point (at 7:00): there's only one node in the network, called the generator, that runs the hashes to create the history. See the whitepaper Section 3 and 7.1.1. The validators then check that history in parallel on a gpu, as you currently pointed out. The system is still decentralized though, as the generator is elected by the validators, as explained in Sections 5.6 and 5.7 of the whitepaper. They also talk about how you could have multiple generators and interlace their respective histories in section 4.4 Horizontal Scaling, however their theoretical limit of 710k TPS doesn't depend on it, as explained in section 7.2 Network Limits. I'd love to hear your new analysis, with this new information in mind!
I’m not an expert but maybe it’s possible to make a loop and appoint in a random intervals one random high tier hash clock validator in the network and make it the only validator that processing the time hashes? regarding the concerns of heavy computing power needed to run such blockchain , I think the tech is moving fast so the current computing power will be much cheaper in two years and so on and therefore will be much broader and much more accessible.. if you guys can consider to add a similar proof of history concept into koinos it can definitely make KOINOS a god tier blockchain due the insanely high speed TPS and a feeless based smart contracts with the many other groundbreaking features! High TPS can accelerate adoption, people love high performance and high numbers ..
I don't understand why they called it "proof of history". That threw me off. It is a proof of work at the "hash level", or "core level" of the network. Like Michael said, it can be a fascinating way of creating systems that rely on a tier model of high probability transactions that propogate upwards until they are "confirmed" at whichever degree that seems appropriate. I suppose if a trasaction fails to confirm, it's not a major disaster. If it's true that they can use these hash ordinals to track "wall clock time", as Michael put it, that opens up a lot of interesting possibilities. "Wall clock time" is only data, after all. With a decentralized micro-clock, entities would be able to theoretically claim ownership of their actions in "real time". For example, as I type these words, the network could secure them as my own, in real-time without me having to "save" or "upload" anything. Node requirements are the only concern. But this seems like only a question of engineering a standard for "probability of trust", rather than a hard limitation. Koinos could definitely leverage this kind of clock.
@@JohnDoe-wq7cp Exactly. In the paper they mention PoH being able to work on top of PoS or PoW, and it seems they chose PoS. But if they took PoW, then what FFMAROTO was suggesting would be true, because the person with the fastest computer would always win the PoW contest and become the PoH leader.
Great explanation! Super interesting! I’d love to see more Koinos explainer videos too
Great explanation from a fantastic team. Bringing class to a Fee-less blockchain 😎
Thank you so much! This is the best explanation I've heard so far.
Hello Michael. I believe your analysis is flawed because you missed one important point (at 7:00): there's only one node in the network, called the generator, that runs the hashes to create the history. See the whitepaper Section 3 and 7.1.1. The validators then check that history in parallel on a gpu, as you currently pointed out. The system is still decentralized though, as the generator is elected by the validators, as explained in Sections 5.6 and 5.7 of the whitepaper.
They also talk about how you could have multiple generators and interlace their respective histories in section 4.4 Horizontal Scaling, however their theoretical limit of 710k TPS doesn't depend on it, as explained in section 7.2 Network Limits.
I'd love to hear your new analysis, with this new information in mind!
Thanks for the explanation. There isn't enough about Solana on the web yet.
Fascinating! Thanks for the good explanation :)
Great vid as always
Best team and nice explanation
I’m not an expert but maybe it’s possible to make a loop and appoint in a random intervals one random high tier hash clock validator in the network and make it the only validator that processing the time hashes?
regarding the concerns of heavy computing power needed to run such blockchain , I think the tech is moving fast so the current computing power will be much cheaper in two years and so on and therefore will be much broader and much more accessible..
if you guys can consider to add a similar proof of history concept into koinos it can definitely make KOINOS a god tier blockchain due the insanely high speed TPS and a feeless based smart contracts with the many other groundbreaking features!
High TPS can accelerate adoption, people love high performance and high numbers ..
Very interesting , thanks for your video
That's really interesting information. Is there another video where you explain what happened in September?
I don't understand why they called it "proof of history". That threw me off.
It is a proof of work at the "hash level", or "core level" of the network.
Like Michael said, it can be a fascinating way of creating systems that rely on a tier model of high probability transactions that propogate upwards until they are "confirmed" at whichever degree that seems appropriate.
I suppose if a trasaction fails to confirm, it's not a major disaster.
If it's true that they can use these hash ordinals to track "wall clock time", as Michael put it, that opens up a lot of interesting possibilities.
"Wall clock time" is only data, after all.
With a decentralized micro-clock, entities would be able to theoretically claim ownership of their actions in "real time".
For example, as I type these words, the network could secure them as my own, in real-time without me having to "save" or "upload" anything.
Node requirements are the only concern. But this seems like only a question of engineering a standard for "probability of trust", rather than a hard limitation.
Koinos could definitely leverage this kind of clock.
👏
what if I have a computer 100 times faster than te other people in the network. Can I take control of the transactions?
In SOL? No I think it's critical to own most of the voting power from Solana Coins
@@JohnDoe-wq7cp Exactly. In the paper they mention PoH being able to work on top of PoS or PoW, and it seems they chose PoS. But if they took PoW, then what FFMAROTO was suggesting would be true, because the person with the fastest computer would always win the PoW contest and become the PoH leader.