Where's Casper, The Friendly Ghost?
A lot has been written about Casper, the upcoming Proof of Stake consensus mechanism for Ethereum. And like in the famous cartoons and movies it's really a "GHOST" because it is derived from the "Greedy Heaviest-Observed Sub- Tree" protocol. Casper has been announced as early as August 2015 - only about a month or so after the Ethereum main network went life - and design work on it started even earlier (in September 2014). So what is it about?
After Stuart Haber and W. Scott Stornetta established the idea of a blockchain in 1991 - and published a weekly hash of their blockchain for verification purposes in the New York Times starting in 1995 - the so far biggest and most widespread application of blockchain is Bitcoin. Bitcoin, as well as many other blockchains - also Ethereum - currently use a mechanism called "Proof of Work" (PoW) to decide about which block (with which content) should become the next block of the respective chain.
The idea was to create security and prevent tampering by distributing the consensus finding in a way where everybody could join and would be rewarded for their participation - some sort of democratization of the blockchain. In the beginning this "mining" - new coins are "created" in the process and put into circulation - really worked for everyone, but due to the - automatically - adapting complexity of the computational "riddles" to be solved (because so many people took part) - the constant "armament" lead to a situation where only miners with highly specialized mining equipment in areas where electric power is cheap have a chance of making profit anymore.
Apart from the negative effect of more centralized mining the real problem created by this is the huge power consumption for PoW at scale. The estimated consumption of all Bitcoin mining world-wide is equal to the power consumption of whole nation states like Ireland. One might argue that mining could be done by only using excessive, renewable energy. But in fact it's done in many different places - without any consideration of how the electricity used was produced. Worse yet, blockchain is a relatively young technology and if we look at the numbers of transactions for Bitcoin and compare it to the number of transactions in e.g. the VISA credit card network, then Bitcoin uses far more energy and only processes a small fraction of the transactions. If we think of scaling Bitcoin to a point where everyone would use it - and even if we consider optimizations like the Lightning Network (LN) - the environmental footprint of Bitcoin long-term seems unbearable. And at this point we are ignoring that there are more than thousand other blockchains out there - most of them also using PoW.
This isn't meant to say that the energy consumption so far has been a complete waste. New technologies need to develop and blockchain was such a new technology. But now that it's maturing, we also need to find a way to make it environment-friendly and sustainable. You don't need to be a hardcore environmentalist to see this. It's like in the automotive industry, where more and more cars use batteries or fuel cells with zero CO2 emissions to replace traditional hydrocarbon-fueled engines: We can't continue without looking for opportunities to bring down energy consumption for blockchains.
And this is where other consensus mechanisms come into play. With a lot of them being out there - like Proof of Burn (PoB), Proof of Authority (PoA), Proof of Elapsed Time (PoET) and Proof of Capacity (PoC) - the second most important after PoW is Proof of Stake (PoS).
Up to now PoS' biggest disadvantage was that it is prone to malicious behavior from participants. But with Casper - which is a slightly different PoS consensus mechanism than the rest - penalties get introduced to discourage participants from manipulating the outcome. If stakeholders bet on an outcome that is highly unlikely (which crystalizes after several rounds of betting) then they risk losing their stake - and hence money. With Casper you neither need big datacenters full of mining equipment nor much energy. Instead you need to possess a certain amount of virtual coins and stake them in order to participate. And stakeholders get rewarded for this participation as well.
But Casper has also other advantages. The danger of so called 51%-attacks - especially for smaller blockchains with PoW and a limited number of miners - is real. Casper on the other hand lets only stakeholders vote and therefore prevents "outsiders" - without any stake in the blockchain - from participating, meaning they can't buy-in by just procuring and operating mining equipment. And to assume that somebody might be able to buy more than 50% of all Ethereum coins is not only unlikely (with a market cap of roughly 30 bn US$ at the time of writing), but due to the mechanism with those betting rounds a successful attack is even harder and the malicious stakeholder would risk losing all his money.
And if one day the consensus mechanism for private or consortium blockchains - currently usually PoA - needs a replacement, than Casper - with its penalty mechanism - might be just the right answer.
So when can we expect Casper to arrive? It has been postponed quite a few times, and this year it was planned to be introduced in February with Ethereum's Constantinople upgrade. Alas, this didn't happen - it was delayed once again. And it won't come all of a sudden. The plan is to introduce it cautiously with only every 100th block being agreed on with PoS/Casper at first - the rest still being PoW -and then increase the number of Casper blocks slowly. Now, we obviously have to wait a bit longer - hoping that it finally will arrive and lead the way to more efficient, secure and environmentally sustainable blockchains. And let's hope that other blockchains will follow soon.