How secure are Proof-Of-Stake systems, how fast they are, and what is wrong with (de)-centralized social media. These questions are asked by many. Let’s try to figure it out together!
Tron vs. Steem (it), or local usurpation
More precisely, Tron is against Steem, but the conflict began with Steemit. Justin Sun bought a stake in Steemit (which is essentially a Dapp on the Steem blockchain) and then tried to move the entire Steem to the Tron blockchain. The community didn’t like it — the Hive hard fork took place.
Who’s guilty? Both sides. The Steem community turned out to be tight-knit, but rather, lazy. It only woke up from a thirtyfold increase in staking value when it played a trick on it. San climbed into a strange monastery with his charter and even attracted some exchanges, which managed to use users’ coins to vote.
This suggests that in (D/L) PoS, the problem of seizing power, known since the days of cooperatives and barbaric raiding, has not been solved. In PoW, for example, the centralization of pools since 2018 is only visual. Nominally, the TOP pools are located in China, but the location of nodes and communities of miners shows that there is physical equipment in Siberia, and Chile, and Venezuela, and Canada, and the United States. The mining leader has also changed: China, Russia, the USA …
What to do? One potential solution that I have been able to find is Fair DPoS, in which voting for delegates assumes that the stake (stake) of the participant is divided by the number of votes. For example, a participant with a share of 100 tokens voted for delegates A and B. The blockchain will take into account that each received 50. If he voted for three, then each will receive 33.33. This makes it possible to neutralize the ability of a large holder to put the necessary delegates in the TOP, and also eliminates corruption.
Anyone who talks about the advantages of (D/L) PoS over PoW and forgets about this aspect is lying at best, and at worst trying to deceive neophytes who are ready to invest and receive their “legitimate” 5% of monthly payments. Not to mention the fact that the holder may turn out to be a fraud at all. This case is still poorly researched — even when coins are frozen for 10–13 weeks, the question remains open.
In short, there is a risk of losing everything because of the management system, even if it seems that the change of ownership does not mean anything.
EOS is a project that killed itself
I like Bitshares and Graphene, but EOS showed from the very beginning how it is possible, but you do not need to earn material capital while losing social capital.
In essence, the longest ICO in history was an almost endless (about a year) token sale, which worked according to the greed formula: emission -> purchase -> price rise -> emission -> price decline -> buy -> new circle. Block.one managed to raise more than $ 4 billion (and then fell out of favor with the SEC).
EOS has other disadvantages as well. First, very expensive “energy”. One developer once told me, “We plan [for this reason] to move to WAX.” The strangest thing is that you can’t pay for a transaction: buy and buy energy! Besides, in EOS, you can, albeit conditionally, but change smart contracts. The level of trust in those is extremely low. Among other cons: no NFT standard, cut down free nodes, no convenient SDK for unit tests, etc.
Of course, all the minuses can be turned into pluses: the same replacement of smart contracts can be interpreted as simplicity in adding options and testing bugs. You can create multiple operations in a single transaction, or you can clean up the memory and sell it (green).
But the fact remains — the advantages declared during the ICO not only did not eliminate the problems of Ethereum but also gave rise to their own. This looks especially bad against the background of the role of EOS whales.
DASH: why doesn’t PoW / PoS help?
I often hear that a compromise between PoW and PoS is the perfect solution. This approach was adopted by Dash, Zilliqa, Decred, and some other projects. But in reality, everything turned out to be not quite so.
In Dash, 90% of the block reward is split between miners and master node operators. Masternodes are needed for two operations: PrivateSend and InstantSend. The first allows you to specify the required number of shuffles to increase the level of anonymity, and the second — for conditionally instant transactions.
Did something go wrong? The biggest disappointment was the terribly expensive D3 miners, which did not pay off so much. Delayed deliveries rush at the end of 2017 and falling prices played a role.
The architecture may look right, but there is always room for epic failures.
If something is sold, it is profitable for someone
The Person of the Year in 2019 was Greta Thunberg, who once again drew the world’s attention to the problem of global warming, although greening has long been in trend. Masternode vendors of various kinds (D/L) PoS systems are taking advantage of the situation by promoting the claim that mining is evil for the planet. They are silent about three facts:
- The heat from many mining farms is used to heat greenhouses, houses, shopping centers, to generate new energy. The detrimental effect of mining is often overstated. In Ethiopia alone, Hidasa (HPP) at 6.45 GW exceeds all Bitcoin mining (just over 5.5 GW).
- The architecture of (D/L) PoS systems does not involve a large number of participants. In PoS systems, 1000 already seems big. When scaling up, given the network effect and Metcalfe’s Law, it is likely that hardware will undergo the same evolution as in traditional mining. Nobody canceled the problem of ping (due to the geo-referencing of the same nodes), network bandwidth, and other characteristics that can be used at the software level only up to a certain limit. Just remember how Google and Facebook work. Sharding also hides the possibility of increasing capacities for storage, computing, etc.
- PoS has so many modifications: DPoS, LPoS, PoS, POI, etc. The difference in the rights, opportunities, responsibilities of various network participants: in the event of the outbreak of war, “iron” will win only the rich.
If you forget about the standard work from the Bitcoin whitepaper and move on to real-life calculations, you need GPU power. And you can’t store data on centralized servers or in the Amazon cloud — here Proof-of-Spacetime and Proof-of-Replication come to the rescue, which is used, for example, in FileCoin, Storj, Sia. As a result, we still return to PoW, albeit in a different form.
Eco-friendly marketing moves are not the only ones in the arsenal of salespeople (D/L) PoS systems.
For example, let’s take the material of A. Kolyayev: voting in ⅔ in Tendermint is a very effective process as long as there are no discrepancies. With its help, you can not only “extract” blocks, but also approve any value/work/result. This achieves the desired level of security, but at the same time creates accessibility problems.
Could it be that ⅔ is not enough? Quite, which means simple and anticipation. It’s good that after approval, the result is fixed as final, but Ethereum has uncle-blocks — an additional bonus for all miners.
In my opinion, the main problem is how the specific gravity of the node that the algorithm will choose is determined. In short, this is the number of coins locked. What is the problem? So this is a bank! Who is richer, that and slippers.
And one more thing: 1000 nodes are somehow not enough for the world. In Bitcoin, where the nodes are not paid for anything, there are about 10,000 of them; in Ethereum — more than 7,000. And this is not bad, since there is no upper limit. But in 1000? The number of 1000 nodes doesn’t sound like an impressive number, but another level of manipulation — yes.
The same problem can be affected by validators: the same ones that were born through IBC. I will quote: “if you send an IBC transaction from blockchain A to blockchain B through a hub, you must trust the validators (nodes that participate in the consensus) of blockchain A, hub, and blockchain B.” Many believe, and not without reason, that this approach can be read as “learning to pay bribes.”
It turns out a rather diverse and high-tech garden: we have nodes, validators, snapshots, and “fishermen” with challenges, which can ultimately lead to chain rollback. That is, what they left from — a chain with great proof of work — they came to that. But then what is the value of the approach?
Especially a lot of questions arise when there are no atomic swaps between PoW blockchains, which are synchronized through IBC, Polkadot, Ethereum, and at the same time, there is a rollback to the desired block within the selected multiblock chain connecting them. So far, I have not seen calculation models where all the costs and associated safety problems will become clear.
Why do we need it?
Sometimes I ask myself the same question, but the answer is still there. Speed. Those Tps.
As CTO The Power noted, “From a technical point of view, all PoS flavors are aimed at gaining speed relative to classic systems. Besides PoS, there are other solutions. PoS was just the first, most likely as a surface option. But the first way that comes to mind is not always the best one. This is what we are seeing now when PoS platforms are passing the test of time. “
There is another misconception in this race. Visa/Mastercard does not have any 20–50 thousand Tps. Real indicators are from 3500 to 4000 Tps. XRP Ledger from Ripple makes 1500 Tps. Zilliqa has excellent performance, as well as Tera and Kadena PoW blockchains.
But 100 thousand and even 1 million Tps may very soon be needed: why?
- decentralized social networks (DSN).
Here the position of the developer A. Piskunov is close to me: “What is a real DSS? First of all, direct subscription. Everyone sees only what they are subscribed to. And they see you only when you create (initialize) the corresponding transaction. In fact, this is what the network of bitcoin and other open solutions is built on.”
This approach requires more balanced decisions and diligence since everyone is tasked with expanding the circle of acquaintances on their own, but you can use triggers: a circle of interests, some learning algorithms (wrote “hello world” — follow on), etc., which is optional, but a necessary condition for evolution. Nobody cancels the subscription to tags/topics either.
Why is the approach good? First, it automatically filters out spam. If not signed, spam will not be sent (today it is enough to know the address/nickname). Secondly, “you see only what you are looking at”: no simulacra! Of course, no one forbids the introduction of recommendations, but they will be based on the user’s self-education, not marketing.
Only this approach can solve the age-old problems of our community:
- My (private) key = my money!
- My assets are the best hedge fund!
- They don’t need intermediaries — they need me, so they can be data oracles, but if I want to and I will validate.
And here we are slowly but surely approaching the main problem that overshadows all the others: it has long been not technical (although this is also), but social. We trust banks, search engines, social networks, and thus we pay them three times:
- Paying taxes / fees / inflation / exchange rate differences. The state creates rules (laws and other regulations) that ultimately help corporations. Even the GDPR looks like the protection of citizens, but in fact, turns us into digital slaves of services. I will not say anything about the Yarovaya package in the Russian Federation or the Snowden / Assange data for the USA, China, and other countries. We are also managed at our expense.
- The rule “if you don’t buy any product, you are the product” works flawlessly. The data is eavesdropped by Siri/Alisa/Google Assistant, it is stored through many .js scripts on websites. Oil of the digital age — big data. And it does not belong to us, although every drop is us. You can (and sometimes need to) sell your data, but you also need to make a profit from it.
- We pay directly by buying advertising campaigns, placing ads, purchasing subscriptions.
In doing so, Facebook blocks ads for cryptocurrencies and then creates Libra. A tax-based RKN can block almost any site and bring trouble to businesses.
So I started researching (D/L) PoS systems. Many DSS prototypes have been made on their basis, but it is stupid to say that they have success ahead of them. Product development, as the ICO boom has shown, is more than just a good team and the right idea.