Emin Gün Sirer talks about the 3 different types of consensus protocols, including an exciting new one released by an anonymous group last year called Avalanche
Watch here:
---
---
Interesting video. Helped me understand avalanche a little more. One confusing thing though, is he said a few times "there is no mining." I think he means, no mining to validate TXs.
But the TXs will still get mined into the BCH blockchain, right? Avalanche is just a pre-filter on the TX to try and prevent double spends attempts by "locking-in" the 1st seen TX.
It's not sufficient itself to maintain the chain. Also, it was a little confusing when he was talking about randomly picking red an blue. In the real system, it's not random.
There's a True correct choice, the real 1st seen TX, & it's very unlikely a double spend attempt will win the metastable race. So once you see your TX accepted by avalanche
I heard that too. I believe what Emin meant by no mining is no conventional mining i.e. "brute-force solve a difficult puzzle", it needs far less energy to "interview 5 nodes".
I think they're 2 different things though, Emin is building a new coin from the ground up just using Avalanche, Chris is using avalanche for preconsensus on bch
I believe Pacia and some other devs were talking about basing Avalanche participant selection on PoW over the past ~100 blocks so it's still critical for gaining control of the network
Yes and no. Basically the old benefit of 51% is "few confirmations" accommodating a split for a double-spend. With Avalanche, network control can cause a sway. But remains _not_ split.
Just that centralization or any kind of attack would still rely upon gaining a large majority of hashrate on the network. Lots of people were irked by the idea of PoS Bitcoin
You think he meant that Avalanche would replace hash mining and not just supplement it? I didn't think that was the plan. I would not be in favor of that.
Now that you mentioned it, I think it is just as susceptible to 51% attacks, and such attack may become easier to deploy because participation no longer requires expensive power.
That's my understanding. Satoshi's consensus method relies on whichever chain is the longest, backed by hashing power. Avalanche was depicted as needing no such "enormous energy".