Seele - New Ideas for Blockchain

Seele's WP is an interesting material that puts into perspective some things that comprise key blockchain mechanics, like distributed consensus, or various aspects of the interaction between the blockchain and external data sources, or addressing resources on the blockchain. While the project's concepts are somewhat vague, they give a reason to have some sort of a conversation about underlying blockchain concepts in general.
So to list Seele project's key features: it's an advanced consensus mechanism (so-called Neural Consensus), then it's an advanced blockchain structure (so-called Heterogeneous Forest Architecture), plus special protocols and naming systems for addressing blockchain resources, as opposed to the incomprehensible ungodly appearance of standard wallet addresses.

Let's start with Neural Consensus Algorithm. This is an idea that claims to surpass existing and coming into existence concepts of Proof-of-work and Proof-of-stake algorithms. Proof-of-work is a process when a special group of blockchain network users, so-called miners, utilizing powerful computing resources, confirms bunches of transactions sealing them with a special kind of signature, block hash, which most important characteristic in this context is that it's very difficult to calculate. The implication of this is that, for example, if I wanted to commit a fraudulent transaction I'd need a computing power exceeding that of all the honest miners bunched together. In other words, I'd need to organize my own coterie of malicious miners, and then, combining our efforts, we would be able to overcome honest miners and calculate hashes for fraudulent transactions quicker than honest miners can calculate hashes for correct transactions (which renders a blockchain branch, containing fraudulent transaction, invalid). Theoretically, it's possible to do that, but there's a very little economic incentive for implementing this scheme since the payoff will never justify all the resources and efforts needed to perform such an attack. Therefore Proof-of-work algorithm is considered a key foundation of the overall reliability of the blockchain. The problem is, it's really costly. Heavy computations necessary to produce hashes require a lot of computing equipment consuming a lot of electricity, which leads, in turn, to environmental problems and so on. A less expensive way to reach blockchain consensus is Proof-of-Stake, a concept where the correctness of transactions is confirmed by a majority vote. It has a lot to do with Game Theory; participants stake cryptocurrency on their votes and in the case when their votes go against the consensus they lose their stakes. When everybody involved doesn't know how the others are going to vote, they'll naturally assume that the majority will vote for blocks with correct transactions and dismiss blocks with fraudulent transactions. To think otherwise would require knowing other's intentions in advance; in this case assuming honest behavior is the safest bet. Therefore Proof-of-stake is considered a reliable mechanism for transaction confirmation without overheads of Proof-of-work. Seele project proposes a new consensus protocol that surpasses both Proof-of-work and Proof-of-stake in terms of potential network scalability. A key argument is that, for example, in PoS voting is a discrete process that takes time to be accomplished. The more nodes are in the network, the more time it takes for all the nodes to reach a consensus. So with increasing network size, block confirmation is going to get slower and slower. On the other hand, the proposed Neural Consensus Algorithm treats voting process not as a discrete but as a continuous process. Every additional vote, being passed to a special function, defines the probability of the eventual outcome. Meanwhile, the network doesn't wait until all of its members decide on the correctness of a current block, it moves on with the voting process continuing at the same time. The correctness of the final outcome is ensured by a special mechanism of sorting blocks in a way that reflects a current voting result.

Another Seele project's key proposition is Heterogeneous Forest Architecture. It addresses the general problem of coexistence of multiple applications on a single blockchain. For example, one episode that brought into the spotlight this problem was the success of CryptoKitties game, which caused near collapse of the Ethereum network due to the influx of users and a sharp increase in transaction activity. The situation raised a question, should all the different apps really share the same space and resources on the blockchain, or maybe it makes more sense for each app to be forked into a separate sub-chain catering specifically to the needs of this app and its users? Another argument, supporting this opinion, is that different apps might require variations in core blockchain architecture, and the current set of blockchain rules and principles may not suit them all equally well. On the other hand, it would lead to the emergence of a multitude of various architecturally different blockchains, each of them not having enough nodes to be sufficiently reliable and robust. Seele project's Heterogeneous Forest Architecture offers to solve this contradiction by introducing a structure where multiple differently organized blockchains are united through a special meta-chain that ensures their interoperability despite the differences in protocols, rules, and underlying architectures. Similarly to how Internet protocols allow very different devices from mainframes to smart toasters to function as a single homogeneous network.
Speaking of Internet protocols, another concept Seele project proposes is a system for resource naming on the blockchain network, or rather, according to the project's vision, on the whole constellation of various interconnected blockchain networks. The system is supposed to be similar to DNS addresses, in other words, comprehensible and user-friendly. It makes sense since the existing practice of using machine-generated unique wallet addresses is pretty inconvenient and in many cases increases the risk of incorrect transactions. In fact, Seele project offers a whole bunch of ideas, regarding the protocols used for communication between blockchain nodes. These concepts present alternatives to currently used transport level Internet protocols, TCP/UDP, for example in the form of VTP (Value Transport Protocol) a transport protocol purportedly designed to better suit the specifics of blockchain node communication. Similarly, Seele project offers VHTTP, a blockchain version of HTTP protocol for the application layer of blockchain communication. VHTTP is also a potential way to organize communication between blockchain resources and standard Internet resources addressed through URIs. This, in turn, leads to an idea of integrated infrastructure, containing both blockchain and non-blockchain resources, addressable in a unified way.

Additionally, Seele project offers technical ideas related to organizing systems, including both on-chain and off-chain data resources. Currently, due to the costs and inefficiency of storing data directly on the blockchain, many projects store data separately, most often using similarly decentralized file storage/data storage solutions such as IPFS or IPDB. Seele project adds a concept of Resource Description Framework (RDF) as a link between blockchain and off-chain stored resources.
Overall, Seele project offers a number of interesting technical concepts, part of which has technical implementations. Experimenting with different potential technological solutions to key blockchain problems either can bring us closer to the viable blockchain architecture or at least to give us a better understanding of those problems and their underlying causes.

---

Useful links

Website | ANN | Whitepaper | Telegram | Facebook | Twitter | Medium | LinkedIn | Bounty

Authored by: faragly (@popeye_the_sailor)