DagCoin: a cryptocurrency without blocks

[monsterer]

Wrong; blocks need timestamps in order to retarget the PoW difficulty.
Quoting from

https://en.bitcoin.it/wiki/Difficulty

"The difficulty is adjusted every 2016 blocks based on the time it took to find the previous 2016 blocks"

Blocks can lie only a little about their timestamp; if too far out of whack, the block
is not relayed.

Notice how the interval is large enough to make the massive errors in timestamp values insignificant? There is something like a 22 minute window of validity for timestamps in bitcoin. They're basically useless, except when dealing with them over a massive interval, such as in the retargeting.

[1715244978]

1715244978

[1715244978]

1715244978

[1715244978]

1715244978

[Come-from-Beyond]

There have to be nodes. There have to be people listening for and forwarding transactions. Maintaining an internal model of the state of the network and forwarding its observations to those who request it.

Right, but consensus model doesn't require to assume anything about nodes.

[Come-from-Beyond]

Wrong; blocks need timestamps in order to retarget the PoW difficulty.

Only if you need supply inflation schedule tied to the outer world timeline, like 150 BTC per hour goal.

[jonald_fyookball]

but whats the advantage of not having blocks?

[Come-from-Beyond]

but whats the advantage of not having blocks?

To name a few:

No mining centralization
No block limit drama
No hard TPS limit
Relaxed requirement for bandwidth
True micropayments

[crazyearner]

but whats the advantage of not having blocks?

To name a few:

No mining centralization
No block limit drama
No hard TPS limit
Relaxed requirement for bandwidth
True micropayments

Nice few advantages their and can see a lot more coming from this in the future. No doubt will lead the way for crypto and indeed maybe change how future crypto coins are created.

[smooth]

Wrong; blocks need timestamps in order to retarget the PoW difficulty.
Quoting from

https://en.bitcoin.it/wiki/Difficulty

"The difficulty is adjusted every 2016 blocks based on the time it took to find the previous 2016 blocks"

Blocks can lie only a little about their timestamp; if too far out of whack, the block
is not relayed.

Notice how the interval is large enough to make the massive errors in timestamp values insignificant? There is something like a 22 minute window of validity for timestamps in bitcoin. They're basically useless, except when dealing with them over a massive interval, such as in the retargeting.

The window of validity is quite a bit wider than that, arguably infinitely wide (at least up to the limit of the timestamp data type). By consensus, the timestamp has to be later than the median of the last 11 blocks, which typically is close to an hour in the past. It can be up to two hours in the future (relative to a node's view of "current time") for relay purposes, but the consensus has no maximum timestamp at all.

In normal usage the window of validity is small compared to the retargeting window, but edge cases can be important too.

[DumbFruit]

Could someone explain how this;

No mining centralization
No block limit drama

is compatible with this;

No hard TPS limit

Thanks..

[jl777]

Could someone explain how this;

No mining centralization
No block limit drama

is compatible with this;

No hard TPS limit

Thanks..

the total transaction capacity scales with the total number of nodes.

[DumbFruit]

Could someone explain how this;

No mining centralization
No block limit drama

is compatible with this;

No hard TPS limit

Thanks..

the total transaction capacity scales with the total number of nodes.

Every "node" could be in the same location, even the same machine.

At a certain frequency the software chooses a transaction that references
a high number of parent nodes.

Of course, not every past transaction could be reachable, as users may decide to never reference
certain published transaction. However, the parent selection, with average out-degree 2, and low network
latency, can guarantee that there will be frequent checkpoints referencing almost all previous
transactions.

After a checkpoint is found, the software updates all nodes reachable from the checkpoint with a
forward pointer to this checkpoint. A checkpoint has its own score counter, initially set to zero. When
the update algorithm reaches a checkpoint node, it increments the counter and stops propagating
backward. The score of a transaction is computed as the last stored score in the transaction plus the
score of the pointed checkpoint. Checkpoints are considered as nodes on the DAG, so the same checkpoint
finding algorithm can make checkpoints that refer to other forward checkpoints.

https://bitslog.files.wordpress.com/2015/09/dagcoin-v41.pdf

[jl777]

Could someone explain how this;

No mining centralization
No block limit drama

is compatible with this;

No hard TPS limit

Thanks..

the total transaction capacity scales with the total number of nodes.

Every "node" could be in the same location, even the same machine.

At a certain frequency the software chooses a transaction that references
a high number of parent nodes.

Of course, not every past transaction could be reachable, as users may decide to never reference
certain published transaction. However, the parent selection, with average out-degree 2, and low network
latency, can guarantee that there will be frequent checkpoints referencing almost all previous
transactions.

After a checkpoint is found, the software updates all nodes reachable from the checkpoint with a
forward pointer to this checkpoint. A checkpoint has its own score counter, initially set to zero. When
the update algorithm reaches a checkpoint node, it increments the counter and stops propagating
backward. The score of a transaction is computed as the last stored score in the transaction plus the
score of the pointed checkpoint. Checkpoints are considered as nodes on the DAG, so the same checkpoint
finding algorithm can make checkpoints that refer to other forward checkpoints.

https://bitslog.files.wordpress.com/2015/09/dagcoin-v41.pdf

are you disagreeing with the statement that the more nodes there are (whatever you call them, wherever they are) the more transactions can be processed?

PoW protects from all nodes being in one location

[DumbFruit]

are you disagreeing with the statement that the more nodes there are (whatever you call them, wherever they are) the more transactions can be processed?

PoW protects from all nodes being in one location

Since "nodes"/checkpoints are based on transactions, that's kind of like saying, "are you disagreeing that more transactions means that more transactions can be processed?" Well no, I'm not disagreeing with that.

PoW only protects from all nodes being in one location (centralization) when it is simultaneously coupled with a limit to transactions per second, otherwise one physical location would out-compete others.

[jl777]

are you disagreeing with the statement that the more nodes there are (whatever you call them, wherever they are) the more transactions can be processed?

PoW protects from all nodes being in one location

Since "nodes"/checkpoints are based on transactions, that's kind of like saying, "are you disagreeing that more transactions means that more transactions can be processed?" Well no, I'm not disagreeing with that.

PoW only protects from all nodes being in one location (centralization) when it is simultaneously coupled with a limit to transactions per second, otherwise one physical location would out-compete others.

I dont understand...

Can you explain how one physical location can out-compete the others, when they are all merged into the single txnet?

The cost to check a tx is much less than to do the PoW, so there can be "islands" pumping out lots of tx, but what prevents them from propagating and merging into a single set of transactions?

why does there have to be any limits to transactions per second?

[tromp]

why does there have to be any limits to transactions per second?

To allow full nodes to operate with limited bandwidth, i.e. to allow decentralization.

[Come-from-Beyond]

Could someone explain how this;

No mining centralization
No block limit drama

is compatible with this;

No hard TPS limit

Thanks..

We used to think that bigger blocks increase rate of orphaned blocks and hence it's more profitable to put all miners into a single location (i.e. centralize). For blockless blockchain it's not a case because it's allowable to "mine" on top of old transactions.

[Come-from-Beyond]

To allow full nodes to operate with limited bandwidth, i.e. to allow decentralization.

DAG makes it much easier to handle intensive data flow because of multicast. A single transaction easily fits into MTU and if it's not received by few of the endpoints it's not a problem. Even more, TPS bursts are automatically smoothed, a full node needs majority of the transactions only if it sends or receives money, in all other cases it needs only to verify PoW and rebroadcast transactions to its neighbors (if multicasting is disabled).

[DumbFruit]

We used to think that bigger blocks increase rate of orphaned blocks and hence it's more profitable to put all miners into a single location (i.e. centralize). For blockless blockchain it's not a case because it's allowable to "mine" on top of old transactions.

We already mine on top of old transactions, that's certainly not novel.

To allow full nodes to operate with limited bandwidth, i.e. to allow decentralization.

Yes.

To allow full nodes to operate with limited bandwidth, i.e. to allow decentralization.

DAG makes it much easier to handle intensive data flow because of multicast. A single transaction easily fits into MTU and if it's not received by few of the endpoints it's not a problem. Even more, TPS bursts are automatically smoothed, a full node needs majority of the transactions only if it sends or receives money, in all other cases it needs only to verify PoW and rebroadcast transactions to its neighbors (if multicasting is disabled).

The full node is the heart of the issue. If there are only a handful of full nodes, and thousands of thin clients, then this protocol isn't decentralized, which is the claim in the whitepaper.

DagCoin is a cryptocurrency design that attempts to be highly decentralized by merging the concepts of transactions and blocks and making each user that transact a miner.

Emphasis mine.

[Come-from-Beyond]

We already mine on top of old transactions

No. We mine on top of recent transactions.

The full node is the heart of the issue. If there are only a handful of full nodes, and thousands of thin clients, then this protocol isn't decentralized, which is the claim in the whitepaper.

Can't comment on that whitepaper.

[DumbFruit]

We already mine on top of old transactions

No. We mine on top of recent transactions.

I was being a little facetious.
I don't see how this protocol avoids the tradeoff between transactions per second and decentralization, and just repeating the claim isn't helping me understand. You mine on old transaction, ok, so what? How does that help us any more than mining on top of recent transactions in regards to centralization?

[Come-from-Beyond]

I was being a little facetious.
I don't see how this protocol avoids the tradeoff between transactions per second and decentralization, and just repeating the claim isn't helping me understand. You mine on old transaction, ok, so what? How does that help us any more than mining on top of recent transactions in regards to centralization?

100 Mbps allows 12'000 TPS for 1 KiB transactions. There is no an incentive to centralize with conventional hardware. VISA peak TPS is reported to be 56'000 TPS (https://en.bitcoin.it/wiki/Scalability). Average is 2'000 TPS.