Node connectivity absolutely matters for mining. Thinking otherwise is retarded.
Discussion
Right. Which is why with DATUM, you run a full node and a miner. Your node broadcasts, your miner hashes. The only thing the pool does (i.e. OCEAN) is the reward splits for valid shares.
The guy in the thread seems to think that the flow is your miner finds a block, it sends the block to OCEAN for some reason, then you broadcast from your node, then somehow, OCEAN invalidates the block because of splits?
He clearly doesn't understand how it works.
I think it's more nuanced than that. It's the question of whether Ocean would broadcast a block submitted as a share. Assuming that Ocean is following the practice of having a well connected (with other miners) node.
OCEAN doesn't broadcast the block... That's the point. The miner submits the block with the shares to OCEAN for pay splits. There are the same number of "hops" because there's no central node hub. A miner can connect to any number of nodes they wish. They propagate the block the same way transaction pools propagate. There isn't some cabal of special nodes that need to be connected to.
There is a cabal of special nodes. They are the mining nodes and as a miner, you want them to build on blocks you broadcast before anyone else's.
How do you become well connected from your home node?
I can't tell if you are trolling. If a mining node is sent a valid block, do you really think they don't start mining with the new merkle hash just incase they could hit two blocks faster than every other miner? Is that really how you think it works? If you can broadcast a transaction to a miner to be included in a block, why as a miner do you think they would disregard a valid block, and by what mechanism? Is this a real theory you have?
There is very clear miscommunication here.
Although it was never answered, I assume Ocean would broadcast a solution via a stratum channel.
My argument is to the latency of propagation for a block as it pertains to nodes and how relevant that latency is for orphan blocks.
Broadcasting from a well-connected mining node will be built on more quickly than a home node broadcasting fan-out from their node.
If a DATUM user submits a solution via stratum, I would hope Ocean broadcasts from their node (assuming it well connected, as is best practice for mining pools)
The latency we are talking about here is less than 500 milliseconds. If you are saying that 500 milliseconds is the deciding factor between a home miner and a commercial mining pool, I think you have a very strange sense of odds and statistics.
Again, a DATUM user broadcasts the transaction not OCEAN. once a valid block is broadcast (and received by the miner nodes as well) they begin to hash the next block.
The scenario you are talking about is unsolvable because of internet communications but also such an infinitesimal chance it's kind of hard to take seriously.
To be clear the scenario is that both a home miner and a corporate pool find blocks simultaneously (already improbable) and because of a more connected node AND the next block being found using the corporate miner's block merkle root. In most cases there are minutes, not milliseconds between valid block broadcasts.
This is such a strange hill to die on especially since it applies to the 2 largest pools as well. Foundry (american based) and AntPool (china based) could have the same thing happen because of the latency discrepancy between the two continents. OCEAN has nothing to do with this problem.
so make sure your node is well connected obviously?
the concern troll you and econo are attempting is embarrassing.
the idea that a pool can exert censorship by leveraging its supposedly superior connectivity is a joke. come back when you're ready to be serious.