Theoretically infinite Transactions per Second Capacity
Peer-to-Peer : All blockchains are peer-to-peer (like ledger on the torrent). Nothing new here
Transactional System : The network can be used for transactions. Nothing new here
Scalability is a problem in the current blockchains.
Current solutions to scalability:
- Faster blocks
- Larger blocks
- Adding supernodes/masternodes to the network dedicated to validating transactions.
- Multiple blockchains
- Side-chains (Lightening network, Raiden, Rootstock, Trinity, Loom, Alpha, Hive, MimbleWimble, etc.
- Sharding (Oversimplified meaning : Not every node stores the whole blockchain and not every node has to verify each transactions.)
Quarkchain also uses sharding to solve the problem of scalability. The way quarkchain is different is that -
- Quarkchain side-chains are elastic (the capacity of the data to be stored on the side chains increases as and when needed)
- Quarkchain side-chains can talk with each other — (In the current implementations of the side-chains, the side chains act as the independent transaction facilitators. Both the sender and receiver have to be on the same chain to complete the transaction). Quarkchain can do the transaction between the side chains. So Alice’s account/address can be on one side-chain and Bob’s account/address can be on the other side-chain and still they will be able to send QuarkCoins to each other. !
- Quarkchain uses horizontal scaling — what is horizontal vs non-horizontal scaling? Lets quickly go over what is a supernode/masternode. Super simplified: Supernodes gets to decide which transactions are valid, who is sending how much money to whom.. and is there double spending happening. To give supernodes this authority to be validator, they need to keep some (large number of ) coins as a collateral. (If they act weird or try to scam — the collateral will be lost forever. ) Supernodes can become so expensive that only a handful people can afford to buy enough coins to create a node (for example Dash masternode requires 1000 Dash which is $460,000!!). To make the blockchain validation as decentralized as possible, quarkchain uses cheap supernodes so a large majority of people can afford the cost of setting up a supernode and become a part of validators. The number of masternodes /supernodes does not dictate the network’s transaction capacity, but the number of shards allowed to proliferate does. So, Personally, I wont call it a horizontal scalability.. but its avoiding centralization of the validation of the network, which increases the security.
- PoW mining security where > 50% mining is dedicated to the main chain so that 51% attacks are avoided. (Not much new in this point, this is true for every single PoW blockchain)
Too many words with negative connotation were used. (why?)
- Out of Control
- Broken out
hmm… The first page with a little scary-ish message. Anyway, that was just first impression, so lets move on.
Next page message: Companies are exploring blockchain.
Main points in the vision:
Ethereum transactions during last 6 months (July 2017 to Jan 2018) have grown by 47x.
Visa claims 56,000 Transactions per second (TPS), Alipay claims 200,000 TPS.
Quarkchain can theoretically scale to infinite scaling with right network implementation. Quarkchain is this groundbreaking innovation that can make it a serious contender for the the highest TPS provider in the decentralized (blockchain, permission-less, no-middleman) and in the centralized (non-blockchain, conventional) solutions.
High TPS will allow for various blockchain based solutions over various sectors of human and tech interactions (almost each and every action in today’s world). Examples: Social media, high-freq trading, IoTs, gaming, and most importantly financial payments.
This section of the whitepaper mainly discusses the current solutions and their pros and cons. If you are a regular blockchain reader, then you know almost all of these issues. But if you are not a “regular” blockchain “technical” reader, then please read the simplified exchanges above these pages. Those descriptions do cover almost all of the content that has been described in here.
Some more points can be made about the security of the blockchain via Proof of work (PoW), Proof of stake (PoS) , hybrid proof of work/stake (PoW/S), multi-chain PoW etc.. but I will cover it later when we discuss the “Collaborative mining” section of this whitepaper.
The important section in this part of the whitepaper, is the trade-off of the solutions for scalability, and privacy.
Essentially, for each of the features of blockchain, whether it is the speed (TPS-capacity), special qualities like (anonymity), or smart contracts, data storage etc. the requirement is the data. Everything needs some “digital space” on the digital ledger (blockchain). To increase the capacity, some decentralization or the security of the network can be traded-off. The benefits of Quarkchain providing that “digital-space”, which can be scaled up-and-down as and when needed far outweigh the trade-off of little departure from network security and almost no compromise of decentralization.
Sidenote (nothing critical here):
The illustration of the “System architecture” of the root chain and shards could have been a lot better. The blue text on the blue background is not clearly readable. I almost could not see the projection of the first block from the “root chain” downwards (which contains the shard chains). Additionally, the background image is completely unnecessary and irrelevant.
Now the interesting and fun part starts about the
Divide and conquer & Collaborative mining PoW
First page describes how does the regular good old linear single blockchain works.
Each block has 2 parts,
- Ledger : Where are all the transactions are stored. The code needed to maintain transactions etc.
- Confirmation/Security part: This part is responsible for cryptographically connect previous block’s both ledger and confirmation part to the new block. The new block needs to be accepted via a signature from PoW/PoS mechanism while avoiding the 51% attack and a possibility of a fork.
The function of 2nd part is critically important
In the Quarkchain, there are shards and a root blockchain.
- Ledger part : The shards are the ledger keepers. If more ledger space is needed, the more number of shards will be created. A root blockchain never stores ledger as its sole function is to validate the transaction. (space on the root chain is too precious to lose it for actually keeping the text transactions. Root blockchain is the “bouncer” in the bar who does not serve customers.. Just a security guard.) These shards are like mini fast blockchains that have their own ledger and confirmation parts .. #Inception_blockchain
- A cryptographic connection between a shard’s confirmation and a root blockchain (which only has confirmation function) is complicated to explain but is one is supposed to simplify it, if the confirmation on the small, faster, shard chain is successful, and the confirmation on the root chain is successful, then and only then the transaction is validated. Otherwise the transactions on that whole shard will be invalidated. Those transactions will be taken care by the other shards or next shard block.
- The shard chains in Quarkchain are 15 times faster so the transaction processing time is 10s. However in reality, the transaction can be valid only if the root chain validates the block, which means the real transaction time should be 150s. Although this confirmation time is quite high (slow confirmations) the capacity of confirmations can compensate for it.
The Collaborative Mining:
This part actually is new and interesting to any PoW blockchain. As discussed earlier, the shards are faster and smaller blockchain with their own confirmation part of the block. Secondly, the root chain has its own confirmation function.
Both of these confirmation functions require some sort of cost mechanism. Quarkchain has chosen PoW cost. Dynamically, 50% of the mining power over the network will be used to mine root chain, while 50% will be used for all the shards together.
Both the main chain and the shard chain have their own reward mechanisms as shown in the page 3 of the Quarkchain network with 18 nodes and 8 shards. It will be interesting to see what mechanisms are involved into choosing which chain to mine on as the difficulty of each chain will be (most likely) determined by the transaction load on that chain. New shards can decrease the load on that sector of the network but current full shard-blocks have to be mined by someone. The reward for such blocks will be higher.
The whitepaper says that there is an “open market” for sellers (selling block rewards) and the buyers (in exchange for the hashing power) within each block, between the shards. I am guessing the miners will have to catch up with the additional coding required to be on the most profitable shard within each block! Professional (hopping) miners will be not only choosing the most profitable blockchain.. now they need to choose the most profitable shard within a blockchain.. #Profit_ception
Additional information for blockchain on the illustration on page 3 is shown here
The height of miner blockchains is about 3800s. So, I am guessing that after 1.055 hrs (3800s) the log of transactions will not be found on the blockchain as the shard chain will not keep the data for all the transactions.
The allocation of the mining power for validation of blocks from both root chain and the shard chains is configurable:
Lets take an example of 50% of the mining power for Root chain and 50% for mining power for shard chains.
Validation on the root chain requires Total_chain_work / 2 (50% of the work)
But Validation on the shard chain requires Total_chain_work/ [2 * number of shards * number of blocks of shard chain in 1 block of root chain (15)]
To attack a shard chain requires half of the validation work for each shard.. but the shard chain attack is guarded by the validation over the root chain.
An 51% attack over the root chain requires Total_chain_work / 2 *51% ~ 25% of the mining power over the network. This is the “trade-off” that was mentioned earlier in the article. The conventional PoW blockchains are protected by requirement of 51% attack but Quarkchain in theory will be vulnerable to 25+% attack.
Now lets go back to the 1st argument of being configurable. If we want to increase the security of root chain, we can allocate all the mining power to the root chain making it essentially like a normal chain (with a huge caveat: THIS CHAIN DOES NOT INCLUDE TRANSACTIONS !! )
So we can allocate 75% mining power to root chain and 25% to shard chains.. The security of the root chain is far critical than the shard chains. Therefore, I expect in early days of implementation, the allocation of the security to root chain will be higher and as the total hashing power increases it will be tiered back to most optimum. I would like the team to post a technical paper with the details on how the risk benefit trade off of different proportions of allocations will be optimized and how will it affect the miners in a simulation scenario.
The anti-centralized horizontal scalability expansion (is mouthful).. but in reality all it means is that the blockchain can validate a transaction on a segment of the network with the availability of relevant information from the older block and current shard-chain confirmations. So as number of shards increases, these shards can use different parts of the network and validate transactions in parallel giving the blockchain as a whole larger ability to process transactions.
Cross shard transactions:
Here we have another interesting issue (with not so clear explanation).
In-shard transaction = both sender and receiver have their addresses on the same small shard chain. Its a simple transaction like regular blockchain.
Cross-shard transaction = sender and receiver are on the separate shards. Now how to confirm this transaction?
My guess is that the root chain confirmation is the only way for the cross talk junction between the shard chains (Quarkchain team or anyone else who has better understanding of this concept, please correct me if I am wrong). Which means a Cross-shard transaction will be solely confirmed by a root chain at the end of each block (which is 15 times slower than the shard chain). This is an interesting side-effect of the sharding that you might be unlucky to have your transaction pending for 15x time than someone else because you are doing a cross-shard transaction. And to add a twist, there is no way to know if you are doing a cross-shard transaction. #Gambler
Cross shard talk
Because the shards can “talk” with each other at the end of each root-chain block, only one address (no matter what shard it is on), can transact on multiple shards. This simplifies the account management as only 1 address is sufficient to transact. In comparison, other sharding solutions require both sender and receiver to be on the same shard, which necessitates creating a new address on a common shards whenever necessary.
In Quarkchain, 1 private key manages, a primary account with an address on the “default” shard and a secondary account contains addresses on rest of the shards. A “smart-wallet” will be able to handle 1000s of addresses on 1000s of shards in the future which seems to be a difficult task to achieve. (but I am not a blockchain savant so I am not really sure how this will work)
Ethereum EVM support
Quarkchain will be able to execute smart contracts via ETH EVM. Whitepaper claims that the dApps on ETH will be able to take advantage of the high TPS throughput of the Quarkchain. This is interesting but seems to be “theoretical possibility” for now.
Edit: After thinking about it and re-reading and talking with the core team members, I came to conclusion that Quarkchain will be able to run ETH-EVM natively on quarkchain itself. So the Ethereum codebase in solidity is directly run over a super high throughput blockchain. So all of the ETH based dApps can be directly deployed on Quarkchain network natively. Enjoy your cryptokitties without worrying about jamming the network!!
A smart wallet will be able to interact with smart contract based tokens over quarkchain. It was never mentioned before.. so I guess not.. but whitepaper needs to be clear about it. Anyway, I will edit this part after talking with the Quarkchain devs.
Edit: As per my previous edits, yes the smart contract support is native to the quarkchain
Sidenote: Why is there an image of fluorescent chromatography, western/southern blot DNA/RNA graph below the text? Its simply distracting!
What gives value to the QKC token?
The token QKC can act as a value carrier only if it has a value. The inherent value to the token can be utilized as a contribution reward. Still the “motivation” to the community part seems unclear. Is this statement in the “contribution reward” about the bounty, staking reward, supernode rewards?
The business plan
Mobile applications = Great
MVP deployment and fast evolution:
I am a bit conflicted. Any blockchain allows deployment of the product on the main chain. Doing it on the main chain just costs money compared to the testnet. How is MVP deployment benefits on Quarkchain are different than other chains?
Demand oriented business:
A strategy of choosing a few businesses for high throughput requirements for testing is a great way to start testing the functionality. However, simultaneously the whitepaper also discusses the “small business dApp” deployment. To a reader like me it seems conflicting. Is the primary goal or priority a few big businesses or both big and small businesses. ( I know I am being very picky but clarity is always appreciated).
IoT and Big data:
AI section of business development are too fuzzy to give any credence. We thought that the IoT market will be a great fit for the projects like IOTA but the implementation even for a chain that does not need much computing power and space to be a node on the network is difficult. I dont really think that a proper blockchain with shards validating are a fit for IoT. The IoT devices may be able to pay for something using a local node on the same network etc etc.. thats why this use case is fuzzy.. Secondly, the AI and big data analytics can be done over any network.
“Quarkchain is open to big data analysts and economists to develop new economic models… to provide feedback to quarckchain for better design” seems to be an academic exercise than a business opportunity. AI and big data analytics can be run on any platform and the optimizations can be derived from it.. Therefore it is not really a Quarkchain business development plan.
Its a good example with very clear benefits to all 3 parties involved in the economic development.
This case requires 1 tokens to be deployed on the quarkchain. I wish the whitepaper could clarify if these tokens are ETH-EVM smart contract tokens cross-talking with Quarkchain? or these are the tokens generated on the Quarkchain natively.
Edit: These tokens are natively generated on the Quarkchain blockchain architecture but are ported features from the Ethereum’s solidity EVM directly.
This is an important distinction because with tokens based on ETH blockchain and cross-talking with Quarkchain, there will still be an impact of congestion on Ethereum blockchain. This will remain as a bottleneck of the throughput.
Edit: These contracts are running on Quarkchain directly so there will not be such bottleneck
Timeline is very simplistic. Not much to discuss there other than the fact that the product will be ready for launch around Q1'19 and we can expect the actual partnerships and deployment of dApps by Q2'19 for streaming profits.
I am amazed by this concept and the whole idea of the project. If successful, the possibilities on Quarkchain are literally unlimited. The biggest bottleneck in current blockchains is the throughput. The whitepaper could be less wordy and vague and more specific. It can definitely be a little bit shorter for anyone to be able to read and follow it in a short amount of time.
The team is promising. I have been to their telegram. The ICO mechanism and the mechanism for airdrop is very sophisticated and the team is trying to do their best for making it sure that everyone gets a small chunk of the tokens. The professionalism in these arrangements and an experienced team shows the dedication and attention to the details.
I wish best to the team for the project.
If any of the team members end up seeing this review of the whitepaper, please answer the questions written here. I am open for continued discussion in the comments. Just hit me up.
I will be deploying a scoring system for whitepapers that I review in the future. Please let me know the projects that you would like me to review “Page by Page” style.
WHITE PAPER RATING: 8.3
1 = Blatant scam. Avoid at all costs!
2 = Avoid this as well, too many red flags.
3 = A terrible project but with some glimmer of hope.
4 = Not recommended, unless you have spare change.
5 = We like the concept, don't have complete faith in the project.
6 = Perhaps good as a short term hold until it hits an exchange.
7 = A good buy, but some concerns.
8 = Strongly consider investing in this.
9 = Potentially great, but we have some reservations.
10 = Sell the wife and kids, this is a home banker!
[more details will be added on how did I reach 8.3 score.. Thank for patience ]
Credits: The ICO scoring system is borrowed from the Orphan blocks’s medium page :)
General review of the project
- Working Testnet ahead of schedule
- Strong team
- Strong advisors
- Solidity integration making it easier to directly port apps from ETH to Quark
- Large number of tokens
- The distribution of tokens skewed towards the team
- Lack of separate technical and general whitepaper
- Scalability is the core problem.
- If beats Zilliqa to the punch, that will be great
- Many blockchains aiming for scalability
- May be a little late to the party