In this post, I will explain the basic about Bitcoin Blockchain and Proof of Work, an algorithm that validate transactions and create new blocks (mining), as detailed in Bitcoin whitepaper, by Satoshi Nakamoto. This apply also for Ethereum and other similar blockchains that use Proof of Work (PoW).

Please feel free to point out any mistakes, as I had some technical and English doubts while writing it

First, we need to understand the concept of hashing, only then we can analyze how blockchain works.

What is a Hash?

A hash is a cryptographic function that receive an input of any length and generate an output of a fixed length , an alphanumerical sequence. Simply put, it is a bunch of random numbers and letters.
This is a great website, it shows how hashing and mining works. You can enter some text there to see how it works.
https://anders.com/blockchain/hash.html
It is important to note that a hash has a fixed size, no matter how big or small the input is. Any small change in one letter in the input will generate a completely different hash (output). This is important if you have a big data, because all this data can be converted to a small hash, which will work as a fingerprint of that data.

Blockchain

The idea proposed by Satoshi is to register transactions on a block. When the proof of work solution for that block is found, it is hashed. Each block includes the previous block’s hash, forming a chain of blocks (blockchain), and each additional block reinforcing the ones before it.



Because of that process, if any transaction in a past block is changed the output hash will be different.
When you make a Bitcoin transaction and you see that you have 5 confirmations, that means that 5 blocks were mined after that transaction. A transaction with 6 confirmations is considered irreversible.

Proof of Work

In the Blockchain, all transactions are publicly announced to all nodes, as there is no central authority who validate transactions.

Proof of Work is an algorithm that rewards participants who solve a cryptographic puzzle in order to validate transactions and create new blocks (this process in known as mining).
This puzzle consists in scanning for a value (called Nonce) that when hashed together with thr block data, the hash begins with a number of zeros.



In this website you can see how this process works. You can try to find a Nonce that when hashed generates a hash that start with 0000 on your own. You will see that it is hard, unless you are very lucky. And you are going to be lucky twice in row, if you do go to a casino or something like that lol.

The difficult in finding that nonce increases exponentially as more zeros are required by the block. Difficult is adjusted every 2016 blocks, based on the average hashrate of those past 2016 blocks, so that 2016 blocks are mined every two weeks. (So a block is mined every 10 minutes)

Miners only get their reward if they are mining the longest chain, as it has the greatest processing power invested. If the honest nodes control the majority processing power, the honest chain will always be the longest and will outpace attacker’s chain.

To modify a past transaction the attacker would have to redo the PoW of the block and all subsequent blocks, and outpace the honest chain. For this to happen, the attacker’s would need more processing power than the honest chain, also known as 51% attack.



Source: bitcoin whitepaper. Images from: https://nakamotoinstitute.org/bitcoin/

Topic also available in Portuguese (short version) and Arabic (by Zwei)

Thank you for taking the time to explain this process. I found your explanation to be help. You could have jumped straight into the proof of work definition but leading it with simple explanation of how the blockchain plays into it adds much needed clarity.

So basically, the longest chain is whatever Jihan is pointing his machines at? this is how I see it at the moment, unfortunately. We need more competition in mining, im looking forward to the next decade to see what others have to offer. Right now im with Luke when it comes to paying attention to hashrate distribution, it's a problem... but changing the PoW would be a disaster and kicking the can down the road. We either need proper game changing approach to PoW or simply we need more competition for Jihan's monopoly. Again, I think things will look different during the next decade, im not too pessimistic.

Yes, the PoW could generate a seriously problem of centralization in the future, but this still  is not a valid reason to change/update the PoW, we still need a consensus of what type of the algorithm will work better, an example of this is to change the SHA2 to ShA3, invalidating all of the asics, or change the PoW to PoS, which theoretically fishish with the high centralization. But all of these examples has their problem and nobody has certainty if these changes can really improve the bitcoin.

I don't think a switch to PoS would decrease the centralization.

This would only 'move' the centralization from people (with lots of money) who buy miner and electricity to people (with also tons of money) who stake their coins.
It might have a small influence on who is mining 'more' blocks. But overall this is not a 'solution' to the 'centralization problem'.

Can you explain about 6 confirmations needed for a transaction to be finalized?

Thanks OP for such nice summary ...

While reading about Openpgp protocol[1] it came to my mind ...Regards to blockchain itself (think in terms of data structure) ..

 ... is the blockchain "data at rest" or "data in transit" ?



[1]


https://gnupg.org/ftp/people/neal/an-advanced-introduction-to-gnupg/openpgp/openpgp.org

I'm not trying to say that will decrease the centralization, but "theoretically", according the most of paper it will (like any other paper about Proof of Storage or DAG say that they are the best protocol).


There is no reason for the number '6', its only a number of confirmation adopted by most of the services, because when you have more confirmations, more difficult to the miners reverse the transaction. In truth its very difficult to happen, the attacker needs a big percentage of the total hash rate to reverse the transaction even after only one or two cofirmations, and try to mine again a block to try to build a new longest chain above the old block is not profitable since you are competing with other miners which are already mining the longest chain.

You can calcule the probability for a miner to reverse a transaction here: people.xiph.org/~greg/attack_success.html

Got it, thank you! However, when I check the tool, with 40% hash-power a group can have more than 50% probability to win over the network. It is not really needed to have enough 51% I guess.

Sorry if this is a very dumb question, but can you help me understand the "longest chain" is the valid chain concept? This is something I'm confused about when it comes to bcash, as it is currently the longer chain. Is it because these are two separate coins at this point? Or does the "Longest chain" refer to mined blocks that become orphaned?