TODO: Video

Consensus is the mechanism used to reach an agreement on the state of a blockchain. For example, when one node maliciously changes something on the block but the others don’t then the majority will rule and ignore or even kick that node out of the network.

A consensus protocol in a blockchain or decentralized system can be roughly broken down into two pieces: a chain selection algorithm and a sybil resistance mechanism.

There are some consensus protocols that have more features but we will focus on these two as they are the most common.

Currently, both Bitcoin and Ethereum use a form of consensus called Nakamoto consensus to determine which is the real blockchain. This is a combination of proof of work and longest chain rule.

There are two types of the Sybil resistance mechanisms: proof of work and proof of stake.

The mining feature is what's known as proof of work. ETH and BTC are currently proof of work blockchain that follows Nakamoto consensus. ETH2 will be using proof of stake.

Proof of work is known as a sybil resistance mechanism because it defines a way to figure out who is the block author, i.e. who did the work to find that mine (solved the mathematical problem) and be the author of that block so all the other nodes can verify that it's accurate. Sybil resistance is a blockchain's ability to defend against users creating a large number of pseudo-anonymous identities to gain a disproportionately advantageous influence over the system and in layman's terms it's basically a way for a blockchain to defend against somebody making a bunch of fake blockchains so that they can get more and more rewards.

Proof of work is sybil resistant because mining a single node has to go through a very computationally expensive process. No matter how many pseudo-anonymous accounts you make, each one still has to undergo this very computationally expensive activity of finding the answer to the proof-of-work problem, in our video demonstration in the previous section it was finding a nonce which gives a hash output starting with four zeros but again each blockchain might have a different problem, in fact some of these blockchains make this riddle intentionally hard or intentionally easy to change what's called the block time, the block time is how long it takes between blocks being published (solved) and it's proportional to how hard these algorithms are, so these problems actually can change depending on how long they want the block time to be.

Proof of work also tells us where these transaction fees and these block rewards go to. When we made a transaction we had to talk about gas and a transaction fee. In a proof of work network they're called miners and in the proof of stake network they're called validators.

In this proof of work system, all these nodes are competing against each other to find the answer to the blockchain riddle, all the nodes are trying to get this answer first because the first node to figure out the answer to the blockchain is going to get that transaction fee.

When a node gets paid they actually get paid in two different ways: one is going to be the transaction fee and another piece is going to be the block reward. The gas price we were altering during a transaction is the transaction fee that we're going to pay to these blockchain nodes for including our transaction.

You've probably heard of the Bitcoin halving before, the halving is referring to this block reward getting cut in half and it's supposed to be cut in half roughly every four years.

The block reward is given to these nodes who successfully mined a block from the blockchain protocol itself. This block reward increases the circulating amount of whatever cryptocurrency that is being rewarded. So these nodes are competing against each other to be the first one to find this transaction or equivalently saying, to be the first one to find the answer to this problem so that they can be the ones to win both this block reward and your transaction fee. Some blockchains like Bitcoin for example have a set time when they are no longer going to give out block rewards and the miners or the nodes are only going to get paid from transaction fees.

The gas fee is paid by whoever initialized the transaction, when we got our funds from the faucet, there was some server and somebody else was paying the transaction fee for us. However when we sent ether from one account to another, our account actually paid some transaction fee to send that ETH.

However, proof of work has some drawbacks as well, it costs a lot of electricity because every single node is running as fast as they can to win this race to get the rewards this leads to obviously an environmental impact. Since proof of work and Nakamoto consensus, a lot of other protocols have taken this idea and gone in a different direction with a different Sybil resistance protocol, a lot of them with the intention to be a lot more environmentally friendly and the most popular one right now is proof of stake. There are some chains that are already using this proof-of-stake protocol and that are live and thriving.

Additionally, ethereum has decided to upgrade to ETH2 which will use this proof of stake algorithm as well it'll also have some other features.

In proof of stake there's also a gas fee but it's paid out to validators instead of miners.

Instead of solving a difficult problem, proof of stake nodes put up some collateral that they're going to behave honestly, aka they stake.

In the case of Ethereum 2, nodes put up some Ethers as a stake that they're going to behave honestly in the network and if they misbehave to the network they are going to be removed some of their stake.

This is a very good Sybil resistance mechanism because if you try to create a whole bunch of accounts then each one of those accounts you have to put up some stake and if you misbehave you're going to run the risk of losing all the money that you put up.

In this system miners are actually called validators because they're no longer mining anything, they're just validating other nodes. Unlike proof of work which every node is racing to be the first one to find the block, in proof of stake nodes are actually randomly chosen to propose the new block and then the rest of the validators will validate if that node has proposed the block honestly as we saw in our previous section it's usually very easy for other nodes to verify if a proposal or a transaction is honest.

Proof of stake obviously has some pros and cons as well. Pros are that it is a great Sybil resistance mechanism and a great way to figure out who the author of a block should be, the other pros are that it's way less computationally expensive to figure out the new block because instead of every single node on the network trying to do this only one node needs to do this and then the rest of the nodes just need to validate it.

The cons are that it's usually considered a slightly less decentralized network due to the upfront staking costs it costs to participate. This gets into a little bit of a philosophical battle on how decentralized is decentralized enough.

The general consensus amongst blockchain engineers is that proof of stake is very decentralized and secure, the massive environmental impact improvement is one of the two main reasons why ETH is shifting to ETH2; it reduces the environmental impact by up to 99 percent.

The decentralized network decides that whichever blockchain has the longest chain or the most number of blocks on it is going to be the chain that they use.

This makes a lot of sense because every additional block that a chain is behind it's going to take more and more computation for it to come up, that's why when we saw in our transaction we saw confirmations.

The number of confirmations is the number of additional blocks added on after our transaction went through in a block, so if we see 2 block confirmations it means that the block that our transaction was in has two blocks ahead of it in the longest chain now.

There are two types of attacks that can happen in the blockchain world.

This is when a single node or entity creates a whole bunch of pseudo-anonymous accounts to try to influence a network. Obviously, on Bitcoin and Ethereum this is really really difficult because the user needs to do many computations in proof of work or have a ton of collateral in proof of stake.

A more prevalent attack is what's known as a 51 percent attack. As we saw as part of the consensus protocol these blockchains are going to agree that the longest chain is the one that they're going to go with.

So as long as it matches up with 51 percent of the rest of the network, this means that if you have the longest chain and you own more than 51 percent of the network you can do what's called a fork in the network and bring the whole network onto your now longest chain.

You can see now that blockchains are very democratic whichever blockchain has the most buy-in and is the longest is the blockchain that the whole system is going to corroborate. When nodes produce a new block and add to the longest chain the other nodes will follow this longest chain and add those blocks to their chains.

However if a group of nodes had enough nodes or enough computational power they could essentially be 51% of the network and influence the network in whatever direction that they wanted, this is what's known as a 51% attack and it's happened on blockchains like Ethereum classic which is not Ethereum. This is why the bigger a blockchain network is the more decentralized and the more secure it becomes.

So if you choose run a node as well you are going to increase the security of the network as a whole.

Gas prices can get really high if a lot of people want to send a transaction at the same time, because a block only has a limited amount of block space (can fit a limited amount of transactions) and the nodes can only add so many nodes. So when a lot of people want to use a blockchain the gas price skyrockets.

This is not very scalable because if we want to add more and more people to these blockchains it's going to cost more and more to use the blockchains. Because more people are going to want to get into these blocks this means that there's kind of a ceiling to how many people can use the system because of the financial constraints that will get imposed as gas prices keep rising.

Ethereum 2 is not only tackling the environmental impact of proof of work by switching to proof of stake but they are also implementing this new methodology called sharding and sharding is a solution to the scalability problem, a sharded blockchain really just means that it's going to be a blockchain of blockchains there is a main chain that's going to coordinate everything amongst several chains that hook into this main chain, this means that there's more chains for people to make transactions on, effectively increasing the amount of block space. Sharding can greatly increase the number of transactions on layer 1 of the blockchain.

Rollups are solutions that perform transaction execution outside the main Ethereum chain (layer 1) but post transaction data on layer 1. As transaction data is on layer 1, rollups are secured by layer 1. Inheriting the security properties of layer 1 while performing execution outside of layer 1 is a defining characteristic of rollups.

They're different from side chains because side chains derive their security from their own protocols, rollups derive their security from the base layers, so examples like Arbitrum and Optimism are going to be as secure as Ethereum.

Layer one refers to any base layer blockchain implementation like Bitcoin's layer one and Ethereum's a layer one, these are the base layer blockchain solutions.

Layer two is any application that is added on top of layer one, i.e. on top of a blockchain. Some examples of layer twos are going to be Chainlink, Arbitraum or Optimism Arbitraum and Optimism are very interesting because they are layer twos that also look to solve this scalability issue, they are what's known as rollups and they roll up their transactions into a layer one like Ethereum.