Even if we assume away the problem of data availability it wouldn't be great to have the entirety of the world's monetary transactions depend on a single company building the blocks and publishing zk-proofs. I guess you can say this block generation can be made to be more open, but I think the incentives are that it would be very centralized regardless, no?
It's a similar problem of Spark: even if you could make it be totally trustless it would still be bad for it to be controlled by a single company.
Login to reply
Replies (1)
I mean it depends on what are you trying to achieve, humans seem to be ok with few seconds latency for anything except very hectic metro lines, in which case you are better off with prepaid cards. But for everything else, if Rootstock has 15s blocks, I believe that is enough, and requires no centralized block producers, just publish your transaction, and IF you want to move from one context to another then you have to generate the ZK proof, but that is just the cost of business, you can obviously atomic swap with someone who generates these proofs and provide the liquidity... But note all of that is no different from paying a miner to include a transaction, that is to say it is so open and so competitive you can hardly do any better.
Now when you need instant payments, as in subsecond confirmations, then your options are;
1. Proof of stake sidechains (Stakechain) on top of Rootstock, which is not actually a single company. But while one confirmation is super fast, the more decentralised these get the longer it takes for finality ... But if you want such fast txns, you might be more happy to risk losing some with reorgs. But you still get more confidence than zeroconf.
2. Centralized payment pools, this one gives you the fastest finality possible, but you have the risk of withholding proofs that are necessary to withdraw or the business going down.. here I believe the absolute best you can do is just use small amounts that you are willing to lose, just like your pocket money, and maybe use many providers at once... That basically boils down to something like Ark, but with more programmability of the contract with the server.
So in conclusion, I think we can scale and indeed solve DA, AND have synchronicity which makes interop between contexts much safer than sidechains that go out of sync, but as soon as you require faster confirmations and faster finality, you must start accepting tradeoffs for different amounts ... But worst case scenario (waiting until Rootstock can't be reorged practically) is still orders of magnitudes faster than finality in any fiat system that merchants are used to, hell even cash notes are only finalised when you take them to a bank and see if they are counterfeit or not.
You don't think that is insanely good outcome? I personally played enough with Rootstock and I think their latency is acceptable, and I don't think that LN or Ark give me much better latency for payments I need to be careful about anyways, and for payments that I don't need to be careful about, I am ok just risking it with a temporary cosigner just like I am ok losing the money in my pocket once or twice in my lifetime