4 ms·

There isn't a legitimate reason to restrict block sizes to 1MB. That is the primary reason for the fork. Some claim that anything larger than 1MB will "hurt de

There isn't a legitimate reason to restrict block sizes to 1MB. That is the primary reason for the fork. Some claim that anything larger than 1MB will "hurt decentralization", but many others have shown this to be a very weak argument (sorry, I don't have any links handy). You don't need every Jane Smith running a full node on a rasberry pi to have robust decentralization.

brndnmtthws 9y ago
There is still a plan to activate 2MB blocks, at a later date (which is part of the Segwit2x proposal). I actually don't fully understand the reasons for why the pro-segwit side doesn't want to increase the block size (even beyond 2MB), except that it's a slippery slope: with Bitcoin, every time you change the block size it requires a hard fork. Increasing the block size isn't a long-term solution, because there's always going to be some new ceiling. Eventually increasing the block size will become impractical. As an aside, some coins have a variable block size to deal with this. AFAIK Monero is the only "major" coin which features variable blocks. It's also the only coin that's actually private.
- andrewla 9y ago
  I think ZCash is the only coin that is private in a real sense; Monero is somewhat private, but as I recall, by watching the mempool, it is possible to see the chains being built and to reconstruct the payments. That's all transient data that is not included in the blockchain itself, but for agents interested in tracing future transactions or logging current transactions there's a significant amount of pseudonymous traceability.
  - jstanley 9y ago
    I think this is FUD about Monero. You don't see any more in the mempool than you do in actual blocks. Can you cite any sources?
    - andrewla 9y ago
      This was based on my own assessment of the whitepaper when it first became prominent. The way that transactions were formed was to build on an existing transaction in the mempool and mix the results together with the ring signatures. Looking at the coin now it appears that there's some optional (soon to be mandatory) protocol changes that will mask the output addresses, but I have not looked closer. The Wikipedia article has two papers that draw similar conclusions about the traceability: http://monerolink.com/monerolink.pdf http://monerolink.com/monerolink.pdf https://drive.google.com/file/d/0B7e8g-wJId8md3FYUGF0TlB5NjQ/view https://drive.google.com/file/d/0B7e8g-wJId8md3FYUGF0TlB5NjQ... RingCT claims to conceal the amount, and "stealth addresses" claims to obscure output addresses, but I have not verified these claims because I'm reluctant to spend more time verifying a coin that I am skeptical about.
      - vocatus_gate 9y ago
        I was honestly pretty skeptical about it for a while too, and only recently came around after finally sitting down and doing some reading on it. Currently Monero is the only coin I'd trust for any sort of transaction requiring privacy.
  - vocatus_gate 9y ago
    I think you might be mistaken. Zcash is definitely NOT private for the majority of transactions. You have to opt-in to obfuscated transactions and so few of them are private that the ones that are really stand out. Monero enforces privacy on all transactions by default. Back in the day you could send non-private transactions but during one of the previous hard-forks they removed that ability.
- fosap 9y ago
  I don't understand why bitcoin is afraid of hardforks. Ethereum hard forks like there is no tomorrow, and so far the number of major block-chain-splits is equal, two. The offical implementation not forking does by no means prevent anybody else from hardforking to begin with, Bitcoin cash demonstrates.
ftlio 9y ago
None have shown it to be a weak argument. None. If you could post your links so that HN can see the kind of pseudo-science the Bitcoin community has had to put up with, it would be much appreciated.
- duckingtest 9y ago
  Calling it 'weak' is giving it too much consideration. The only sensible argument against larger blocks is that they boost mining centralization because larger blocks may take more time to propagate. Which may be true but it requires 1GB+ blocks. 10Gbps connection isn't a noticeable cost item for a serious mining operation, which means latency for a 1MB block and a 1GB one should be almost identical. That's it. Remember that decentralization in bitcoin concerns miners, and miners only. 'Nodes' in the whitepaper are synonymous with miners. Non-mining wallets need to have enough bandwidth to download blocks from miners as they appear. For 1GB blocks 20Mbps is enough. Which means a full verifying wallet for 1GB blocks is practical on a LTE connection. Storage is also a non-problem with utxo commitments.
  - ftlio 9y ago
    > Calling it 'weak' is giving it too much consideration You can only say this if you're too ignorant to be worth a damn on the subject, or are insane enough to think that the people who have developed the protocol for the last 8 years are all misguided. If you launch into some conspiracy about Blockstream, Inc., then we'll know it's a combination of both. > Remember that decentralization in bitcoin concerns miners, and miners only And exchanges, merchants, consumers. If you have a stake in the ledger by owning or accepting BTC, then you have a stake in the monetary policy of BTC. If you have a stake in the monetary policy of BTC, then you can only hope to enforce it (or attempt to change it) by running a node. That's before the mathematics of bigger blocks, where the decentralization of miners is threatened even by a 2 MB block size, as those with more hash rate get another bonus due to the latency increases of even a 1 MB block size increase. If all that mattered in Bitcoin was the decentralization of miners, this game would have been over a long time ago. Anybody who thinks that even a landscape of thousands of miners, pooled by very few pools, is decentralized enough on its own, without others validating and auditing the blockchain, is too capable of assumptions to be let near engineering. Why does every currency that has ever existed get devalued into dogshit by its policymakers? Every currency dies because those close to the mint can afford devaluation more than those further from it.

brndnmtthws 9y ago

There is still a plan to activate 2MB blocks, at a later date (which is part of the Segwit2x proposal). I actually don't fully understand the reasons for why the pro-segwit side doesn't want to increase the block size (even beyond 2MB), except that it's a slippery slope: with Bitcoin, every time you change the block size it requires a hard fork. Increasing the block size isn't a long-term solution, because there's always going to be some new ceiling. Eventually increasing the block size will become impractical. As an aside, some coins have a variable block size to deal with this. AFAIK Monero is the only "major" coin which features variable blocks. It's also the only coin that's actually private.

andrewla 9y ago
I think ZCash is the only coin that is private in a real sense; Monero is somewhat private, but as I recall, by watching the mempool, it is possible to see the chains being built and to reconstruct the payments. That's all transient data that is not included in the blockchain itself, but for agents interested in tracing future transactions or logging current transactions there's a significant amount of pseudonymous traceability.
- jstanley 9y ago
  I think this is FUD about Monero. You don't see any more in the mempool than you do in actual blocks. Can you cite any sources?
  - andrewla 9y ago
    This was based on my own assessment of the whitepaper when it first became prominent. The way that transactions were formed was to build on an existing transaction in the mempool and mix the results together with the ring signatures. Looking at the coin now it appears that there's some optional (soon to be mandatory) protocol changes that will mask the output addresses, but I have not looked closer. The Wikipedia article has two papers that draw similar conclusions about the traceability: http://monerolink.com/monerolink.pdf http://monerolink.com/monerolink.pdf https://drive.google.com/file/d/0B7e8g-wJId8md3FYUGF0TlB5NjQ/view https://drive.google.com/file/d/0B7e8g-wJId8md3FYUGF0TlB5NjQ... RingCT claims to conceal the amount, and "stealth addresses" claims to obscure output addresses, but I have not verified these claims because I'm reluctant to spend more time verifying a coin that I am skeptical about.
    - vocatus_gate 9y ago
      I was honestly pretty skeptical about it for a while too, and only recently came around after finally sitting down and doing some reading on it. Currently Monero is the only coin I'd trust for any sort of transaction requiring privacy.
- vocatus_gate 9y ago
  I think you might be mistaken. Zcash is definitely NOT private for the majority of transactions. You have to opt-in to obfuscated transactions and so few of them are private that the ones that are really stand out. Monero enforces privacy on all transactions by default. Back in the day you could send non-private transactions but during one of the previous hard-forks they removed that ability.
fosap 9y ago
I don't understand why bitcoin is afraid of hardforks. Ethereum hard forks like there is no tomorrow, and so far the number of major block-chain-splits is equal, two. The offical implementation not forking does by no means prevent anybody else from hardforking to begin with, Bitcoin cash demonstrates.

ftlio 9y ago

None have shown it to be a weak argument. None. If you could post your links so that HN can see the kind of pseudo-science the Bitcoin community has had to put up with, it would be much appreciated.

duckingtest 9y ago
Calling it 'weak' is giving it too much consideration. The only sensible argument against larger blocks is that they boost mining centralization because larger blocks may take more time to propagate. Which may be true but it requires 1GB+ blocks. 10Gbps connection isn't a noticeable cost item for a serious mining operation, which means latency for a 1MB block and a 1GB one should be almost identical. That's it. Remember that decentralization in bitcoin concerns miners, and miners only. 'Nodes' in the whitepaper are synonymous with miners. Non-mining wallets need to have enough bandwidth to download blocks from miners as they appear. For 1GB blocks 20Mbps is enough. Which means a full verifying wallet for 1GB blocks is practical on a LTE connection. Storage is also a non-problem with utxo commitments.
- ftlio 9y ago
  > Calling it 'weak' is giving it too much consideration You can only say this if you're too ignorant to be worth a damn on the subject, or are insane enough to think that the people who have developed the protocol for the last 8 years are all misguided. If you launch into some conspiracy about Blockstream, Inc., then we'll know it's a combination of both. > Remember that decentralization in bitcoin concerns miners, and miners only And exchanges, merchants, consumers. If you have a stake in the ledger by owning or accepting BTC, then you have a stake in the monetary policy of BTC. If you have a stake in the monetary policy of BTC, then you can only hope to enforce it (or attempt to change it) by running a node. That's before the mathematics of bigger blocks, where the decentralization of miners is threatened even by a 2 MB block size, as those with more hash rate get another bonus due to the latency increases of even a 1 MB block size increase. If all that mattered in Bitcoin was the decentralization of miners, this game would have been over a long time ago. Anybody who thinks that even a landscape of thousands of miners, pooled by very few pools, is decentralized enough on its own, without others validating and auditing the blockchain, is too capable of assumptions to be let near engineering. Why does every currency that has ever existed get devalued into dogshit by its policymakers? Every currency dies because those close to the mint can afford devaluation more than those further from it.