Ah, interesting. I hadn't realised they were operating as a reserve broker, though either way I really don't want them or anyone else encouraging the use of cryptocurrencies.
Conversation
Specifically, I don't want to create a distinction between transactions-within-broker and transactions-between-brokers that further mask away and confuse the impacts that the use of these currencies incur
1
On-chain transactions are only usable as a settlement layer. The fee for sending 1000 BTC is the same as 1 BTC if both were a single UTXO. The arbitrary limits on block size and time don't accommodate doing on-chain transactions for absolutely everything. Have to use 2nd layers.
1
Nothing prevents making Bitcoin support 1000x as many transactions by having 10x faster block time and 100x larger blocks. It isn't a viable way to scale it up to support every little transaction though.
Resource use for mining isn't directly related to transaction throughput.
1
If it supported 1000x as many transactions, it would need 1000x more storage space, so it would be much harder to run a node. There would be more space, so more transactions, but lower fees. Miners would end up getting near zero money from fees until it filled up to 100% again.
2
It would seem this is why Libra (or whatever it's called now) is explicitly architected to allow throwing history away.
1
Even if you can throw away the old history, there's still the issue of bandwidth, latency, etc. It's less common now, but some miners used to produce empty blocks with no transactions simply to get the block rewards due to fees not being a substantial portion of the revenue.
1
1
1
They're only hashing the header, so it may seem strange that they'd throw away the money from the fees. Some of the miners in China were apparently having issues fully syncing with the network due to the Great Firewall, etc. so they'd just make empty blocks instead...
1
1
1
Note: nodes are still validating even without mining. The mining security isn't needed for the basic transaction validation, etc. It's needed because the valid history path is the one with the longest chain of blocks. It's the consensus mechanism for which block(s) are canonical.
2
1
1
Replying to
I'm not… completely sure whether Libra nodes actually get fees. The envisioned funding model was that there would be a tether fund and Facebook or whoever would collect interest on the tether fund and keep it.
1
Replying to
Normal Bitcoin node doesn't get fees, only a mining one. Non-mining node is only relevant to the extent it's being used by users to validate, broadcast transactions, etc.
Can theoretically earn fees from Lightning nodes but fees are so low transactions are essentially all free.
Replying to
I'm not totally certain whether "validators" (the libra equivalent of mining nodes) get fees or not. Even if fees exist I think the idea was supposed to be many validators are run by interested parties (eg payment apps) so when they add a block they just shove their own TXNs on.
1
Replying to
Liquid has 15 parties taking turns in round-robin and then 11/15 need to approve each block. They'll be increasing the number from 15.
help.blockstream.com/hc/en-us/artic
It's widely used by exchanges, etc.
Essentially Bitcoin with trusted parties instead of mining and permissionless P2P.
1
1
2
If you run a Bitcoin node, it just sits there using some bandwidth and a tiny bit of compute resources for validating. It accomplishes nothing unless you use it for something, like pointing a wallet at it or running a Lightning node. Doesn't get fees, etc. for the work it does.


