Since I am still waiting for my internet connection to recover I thought that I could at least write down the basic ideas behind the Multiverse consensus:
Blockchain tracks a single ledger state of a single longest chain.
Blocks are only valid if they contain no double spends and the block producers act as gate keepers who control which transactions are added to the ledger state. Nodes track the "outcome of consensus".
In a DAG based DLT like the one used by IOTA there are no block producers ...
... and anybody can add anything to the underlying data structure at any given time without asking anybody else for permission - even double spends. Every double spend introduces a different version of the ledger state that might or might not be compatible with other ...
... versions introduced by other transactions. IOTA has developed a data structure - the parallel reality based ledger state - that allows us to efficiently track all of these different versions (realities) simultaneously. It tracks the ledger state "before consensus".
Since there can be double spends in the underlying data structure, IOTA similarly to other DAG based DLTs uses a voting mechanism where nodes exchange their opinions to come to an agreement which transactions are supposed to be "accepted" by the network.
Due to our unique ledger state, IOTA in contrast to other DAG based DLTs doesn't have to vote on non-conflicting transactions but only on the conflicting ones which significantly reduces the message overhead.
Multiverse consensus now goes one step further and utilizes the ...
... underlying data structure itself as a voting mechanism. Every transaction that is issued by a node already contains the opinion of that node since it spends certain funds that belong to a certain reality and it also approves 2 or more other transactions in the DAG which ...
... are also part of a certain reality. We therefore already know the opinion of the issuing nodes without having to ask them. This allows us to get rid of the additional voting mechanism which turns the message overhead for voting to 0.
This is equivalent to the ...
"original vision" of IOTA where the only thing that nodes have to do is to issue transactions and consensus is achieved automatically.
To secure the network, nodes with a lot of weight in the consensus process can every now and then send a single transaction that gets ...
... replicated via gossip instead of having to answer to queries of a potentially unbounded number of other nodes. This is not just much more efficient but also much more secure since every node will have the same perception of things and nodes can never ever come to a ...
... different conclusion.
This is essentially the basic idea behind Multiverse and the reason why it is named the way it is.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
If you are a DLT researcher, then you have most probably dreamt of a consensus mechanism:
- without miners
- that doesn't require nodes to query each other
- where nodes "magically" reach consensus in less than a second without wasting network throughput on voting
- that supports an unlimited amount of nodes and validators
- works equally well with a small permissioned set of validators
- that is compatible with PoS, PoW, VDF's and so on ...
- where you can reach finality in 1-2 seconds
- where the chance for an agreement failure is 0
- that doesn't need reattachment or reorganization of the underlying data structure
- where sending a transaction just requires you to gossip your transaction and be done
- where nodes collaboratively validate transactions instead of only the highest weight nodes making decisions
IOTA (much like the universe) doesn't have a total order of events. But in the same way as the universe still allows you to have a total order of events within a "frame of reference" IOTA also allows you to have a total ...
@facemrook@Adrian_Midas@IOTAFanClub@Vrom14286662 ... order of events when it comes to a particular chain of spends. Colored coins allow you to create a traceable and non-forkable chain of spends which creates something like a "frame of reference" in which total order can exist.
Ich möchte mich hiermit öffentlich dafür Entschuldigen dass mein letzter Tweet etwas provokant klang. Ich hatte weder die Absicht @blocktrainer öffentlich anzugreifen noch in irgendeiner Weise bloß zustellen.
Crypto ist ein unglaublich komplexes Thema und man kann nicht ...
... erwarten dass eine einzelne Person über alle Details eines jeden Projektes bescheid weiß. Gerade wenn man über mögliche Auswirkungen von gewissen Unterschieden in der Funktionsweise philosophiert kann es durchaus sein, dass man mal etwas sagt was diskussionswürdig ist.
Roman spricht in seinen Videos frei und ohne Skript und aus eigener Erfahrung weiß ich das Dinge manchmal einfach anders rüber kommen als man sie meint oder man vlt. im Eifer des Gefechts einfach nicht die richtigen Worte findet.
@blocktrainer Ich habe mir gestern mal dein neuestes Video angeguckt und ich muss sagen, dass du da ein paar Sachen erzählt hast die totaler Blödsinn sind.
1. Ein DAG-based DLT bewahrt Daten genauso global auf wie eine Blockchain (inklusiver voller Historie und aller Balances).
Sharding und Snapshotting sind unabhängig davon ob du einen DAG benutzt und das von dir diskutierte NANO hat keins von Beidem.
2. Niemand zahlt in Bitcoin jemals dafür Daten für immer aufzubewahren. Die gesamten Fees gehen an die Miner und nicht an die Nodebetreiber.
Das Netzwerk bezahlt sich also nicht selbst und sorgt dafür dass die Kosten der Nodebetreiber gedeckt sind und in der Tat nutzen Miner ASICS ohne Festplatten, die weder Daten für andere Nodes zur Verfügung stellen noch in irgendeiner Weise die "Infrastruktur" unterstützen.
One that is collaboratively validated by the economic actors of the world (coporations, companies, foundations, states, people) or one that is validated by an anonymous group of wealthy crypto holders?
The problem with current DLTs is that we use protection mechanisms like Proof of Work and Proof of Stake that are inherently hard to shard.
The more shards you have, the more you have to distribute your hashing power and your stake and the less secure the system becomes.
Real world identities (i.e. all the big economic actors) however could shard into as many shards as necessary without making the system less secure.
Todays DLTs waste trust in the same way as PoW wastes energy.
Colored coins are the most misunderstood upcoming feature of the IOTA protocol.
A lot of people see them just as a competitor to ERC-20 tokens on ETH and therefore a way of tokenizing things on IOTA, but they are much more important because they enable "consensus on data".
By creating a single uniquely colored coin that you can spend over and over to yourself, you can create a chain of spends that is secured by consensus.
By submitting data with these spends, everybody in the network will not only agree on the order of the published data but ...
... would even agree on which data to use if conflicting information where published. This is incredibly important for several use cases beyond tokenization.
In fact, IOTAs upcoming smart contracts will heavily use this new ability to build "private unforkable blockchains" ...