NEO consensus algorithm dBFT has been in the works for over half a year. Now according to Eirk Zhang, founder of NEO, the project is near completion.
After half a year of development, the improved NEO consensus algorithm dBFT is about to be completed. dBFT will become the best consensus mechanism for blockchains.— Erik Zhang (@neoerikzhang) January 3, 2019
Delegated Byzantine Fault Tolerance (dBFT) is - apart from a mouthful to say - a consensus algorithm, much like proof of work or proof of stake, which was developed and popularized by the NEO project team. The point of all consensus algorithms is to overcome one quizzical problem that plagues all decentralization: The Byzantine Generals Problem.
This problem gave rise to such consensus algorithms as a way to avoid the Byzantine Generals Problem which is, effectively - and in fairly reductive terms - proof of honesty within a decentralized protocol.
The NEO team developed dBFT as a way to combat this universal problem. The designed system comprised of nodes, delegates, and a speaker and works much like the infrastructure of a country and its government. The citizens, (nodes) vote for delegates (bookkeeping nodes), much like how state representatives are chosen for Congress during elections.
Delegates implement the will of the citizens (ie. keep track of a transaction and record them upon the ledger). Delegates are also chosen, at random, to become the speaker. Speakers verify the transactions on the ledger, creating a block; this then gets sent back delegates of which 2/3rds are required to ensure the speaker's proper verification of said block before it gets added to the network. If less than 1/3rd of delegates agree on the speaker verified block a new speaker gets elected at random and the cycle continues.
Semantics aside, Zhang asserts that implementation of this new consensus algorithm will provide more “strict finality” adding that confirmations will be reduced to just one with an estimation of 15 seconds completion time:
With this improvement, dBFT will have more strict finality. Users only need to wait for one confirmation (15 seconds) to ensure the irreversibility of the transactions and prevent double-spending. This is very suitable for financial applications.— Erik Zhang (@neoerikzhang) January 3, 2019
Zhang stated in a further tweet that NEO wishes to eventually run “large scale commercial applications” highlights the ways in which the project intends to achieve this goal:
The second is to develop a distributed storage network, NeoFS, so that applications can store massive amounts of data. The development of the first release candidate (RC1) of NeoFS is expected to be completed in Q3 2019.— Erik Zhang (@neoerikzhang) January 10, 2019
According to Malcolm Lerider, Senior R&D manager at NEO, DBTF is purpose built with such goals in mind:
"dBFT is invented by NEO and has proved to work well. It is a consensus algorithm developed with perfect finality, meaning that all transactions are 100 percent final after the first confirmation. The blockchain cannot fork with dBFT, and high-value chained transactions are trivial and executed much faster. It is built with regulatory and business use cases in mind." He said.
Zhang received a fair bit of support with this latest development; with many congratulating the months of hard work:
You're a machine. Thanks Erik!— xen (@xenzor1) January 3, 2019
Congrats Erik. We know you've been working hard on this -- and that it's *really* hard to get right.— Mark Jeffrey (@markjeffrey) January 3, 2019
And we (Guardian Circle) would absolutely use NeoFS ... this would be hugely useful.— Mark Jeffrey (@markjeffrey) January 10, 2019
NEO has been working and improving under the whole 2018. I know this is a winner 🤩— Dexter (@DexterToday) January 3, 2019
Read more: NEO organizes 2nd DevCon event in Seattle, 'We keep on going'; Update: NEO's Erik Zhang responds to claims of a security flaw allowing theft; NEO releases decentralization roadmap; Ethereum Classic hit by 51% attack, seeing double spends totaling $460,000