451–460 of 576 posts
Earlier quoted context omitted.
I don't think I have to trust anyone when I run my own Bitcoin node and sign transactions. What entities do you mean?
Don't you need to also trust all of the other nodes on the network to come to consensus?
If majority of the nodes or the miners suddenly conspire to change the rules, my node will simply ignore those blocks as invalid.
Blockchain has no use other than to store tokens for speculative gambling. Take a look at this presentation - https://bit.ly/3OB5G18 Proof-of-work - the USP of blockchain makes it slow and expensive and planet destroying at scale. But it's not a bug, it's a feature. You take it out and the blockchain is neither public not permission-less. A private, trusted, permissioned blockchain is pointless. The whole blockchain,…
I wonder whether one can postulate an impossibility theorem (call it SEV, similar to CAP) for public blockchains (at least using PoW, and whether one can overcome it with PoS or similar is an open question): You can't have a (PoW) public blockchain that is Secure, Efficient (=not planet destroyingly wasteful), and Valuable. Pick any two.
Earlier quoted context omitted.
I agree that databases on their own are never tamper resistant. But you can certainly set up a process around the database that guarantees tamper resistance. You'll still have to trust said process, and with blockchain you still have to trust the code. You'll never get away from having to trust _something_, but at least it is in theory feasible to have tamper resistance both with a centralised database and with a blo…
in practice, there's a difference between "our SREs and management promise to not change anything" and "it would cost x billion dollars to replace this record with a different one, based on previous attacks we've observed"
But you can lock a database down to the point where only external auditors have access to the admin keys. It would make debugging challenging for sure, but that's anyway a tricky part of distributed systems such as bitcoin.
Earlier quoted context omitted.
Your comment goes in all directions so it’s hard to answer. Perhaps it’s useful to point that proof of work is not ubiquitously used in blockchains and much greener alternatives have been preferred in pretty much any crypto that’s been released in the past 5 years
Proof of work is still what powers the vast majority of block chains by value. Perhaps if ETH goes to PoS in some future this will change, but until then, all of the big permission-less crypto curencies are proof of work.
Earlier quoted context omitted.
I wonder whether one can postulate an impossibility theorem (call it SEV, similar to CAP) for public blockchains (at least using PoW, and whether one can overcome it with PoS or similar is an open question): You can't have a (PoW) public blockchain that is Secure, Efficient (=not planet destroyingly wasteful), and Valuable. Pick any two.
Secure and efficient. Money should be a medium of transfer, not a store of value. The fact remains that you can't really get "efficient" with proof of work, and claims about "secure" are dubious with proof of stake.
Earlier quoted context omitted.
How so? With git, you have to manually resolve merge conflicts. With blockchain, the general point is to use PoW (or something equivalent) to automatically do a vote on which is the correct version. Isn't that also the case in a private blockchain?
I think git has: 1. A chain of blocks referencing its parents by hash, thus verifying their validity and immutability 2. A rules for consensus on choosing what blocks can be appended (both technically and organizationally through each company’s PR review process) That makes git a blockchain.
Earlier quoted context omitted.
Why use bitcoin, the slowest blockchain and most inefficient blockchain, as an example? Blockchains like Sui offer basically unlimited transactions/s
"Why use the most popular and most well-known and most used and most heavily-invested-in example of this technology, when I know of this obscure one that promises something that it certainly hasn't demonstrated yet?"
Earlier quoted context omitted.
Why use bitcoin, the slowest blockchain and most inefficient blockchain, as an example? Blockchains like Sui offer basically unlimited transactions/s
Fundamentally the transaction rate is limited by how much data you're willing to store. Bitcoin's 5tx/sec translates to ~5gb/month added to the blockchain - improving upon this requires either making the data stored per transaction smaller, increasing the blockchain growth rate, or moving transactions off-chain.
Earlier quoted context omitted.
Cardano is a cryptocurrency that offers ~250 transactions per second but should soon scale without a layer 2 solution. I mention this as I run a validator on a 8gb raspberry pi with ZRAM swap enabled with no issues consuming ~5W the whole network could run this way if needed. This is lower power due to a mechanism called proof of stake. BTC uses proof of work. Ethereum has been 'trying' to migrate to POS for some tim…
Proof of stake is useless. It just give all the powers to the current stakeholders. So it’s not trustless by definition since you need to trust the stakeholders.
Earlier quoted context omitted.
Your second statement presupposes central authorities can be trusted. Some presuppose that they cannot.
Let me quote from a 2012 interview with Bruce Schneier about his book Liars and Outliers: Q: Given the dangers out there, should we trust anyone? Isn’t “trust no one” the first rule of security? Bruce: It might be the first rule of security, but it’s the worst rule of society. I don’t think I could even total up all the people, institutions and systems I trusted today. I trusted that the gas company would continue to…