Auf Pangolin bin ich wirklich sehr gespannt. Die Frage ist, wie viele Nutzer von Sushi und Uniswap erreicht werden können und wie man diese dazu bekommt ihre Liquidity auf Avalanche bereitzustellen. Da muss beim Marketing echt was passieren. Eine kleine Aufgabe wie gerade im Hub reicht diesbezüglich meiner Meinung nach nicht aus.
Edit: Habe soeben von SEQ eine Antwort erhalten, welche ich euch nicht vorenthalten möchte:
A leader is chosen out of the 1000 pools or so and they add the block to the chain without checking with any other validator before. The next slot a different leader is chosen they can either build on top of the last block to confirm its valid or they can fork.
The protocol basically assumes the recent blocks as not final and not fully trusted and then after an agreed number of blocks in the consensus its deemed as final / settled. So if a block happens every 20 seconds and 15 blocks before it's considered final then the finality is 5 minutes and 14 other validators have effectively confirmed to have validated the block by extending the chain
I haven't looked too much into Cardano if i'm honest. Its on my list at some point to look into in more detail. From my understanding there are multiple types of nodes - core nodes which are block producing and relay nodes (which there can be many more than the equivalent block producing ones). Staking pools will consist of both types and owners / delegators can stake against a pool.
Their consensus is the longest chain like with nakamoto but doesn't have a finality gadget like ETH 2,0 etc. So one leader from a staking pool is selected every slot based upon stake, within an epoch and they produce a block. Others can choose to build on top of that or can fork etc. So its not BFT where every validator has to validate it before it gets added to the chain. This means finality is long like with Bitcoin as transactions can be reverted. They have faster block time than Bitcoin and I have seen figures suggest 5 minutes finality based upon less than 10% of the stake being malicious. (Obviously the time increases if you want assurances against protection of higher % safety threshold.
The important metric isnt the total number of validators, it's the number of validators that verify a transaction before its seen as final. So whilst cardano may be able to scale to lots of validators only a small number (leader for each slot) is confirming before its seen as final
Would make a very interesting chart when only include nodes that participate in a round to validate a block / transaction. The majority of blockchains would be less than 20 nodes, really showing the huge difference and benefits of the avalanche consensus