6 ms·

Arguments in the Bitcoin Block Size Debate

46 points by notsony 11y ago

simonebrunozzi 11y ago
In my view, it's a great and simple short term fix, waiting for more volumes and bigger bucks to get involved. I like it.
itistoday2 11y ago
Relevant: coverage of various perspectives and positions by Ethan Heilman along with counterarguments to each: https://github.com/EthanHeilman/BlockSizeDebate/blob/master/README.md https://github.com/EthanHeilman/BlockSizeDebate/blob/master/... I like that it also includes discussion of the Lightning Network's role, and the idea of making the block size dynamic. These are less heard of perspectives that nevertheless seem significant.
Animats 11y ago
The only people who have a vote on this are the big miners. Unless > 50% of the miners, by hash capacity, make a change that affects the blockchain, it doesn't happen. The top 4 mining operations control more than 50% of the hash rate. Nobody else matters.[1] If the blockchain runs out of capacity, the transactions with the biggest fee attached get processed first. So the free market will solve any overload problem. Bitcoin volume in dollars isn't increasing.[2] The peak was over a year ago. The number of transactions is up, but not the total dollar value. Some of the transactions are just "mixing" transactions, to assist in money laundering. The companies that offer such services are quite open about this.[3] [1] https://blockchain.info/pools https://blockchain.info/pools [2] https://blockchain.info/charts/estimated-transaction-volume-usd https://blockchain.info/charts/estimated-transaction-volume-... [3] https://bitlaunder.com/ https://bitlaunder.com/
- beaner 11y ago
  > So the free market will solve any overload problem. What do you mean by this? If 2mb of transaction data is generated every 10 minutes, but only 1mb of it is mined, what is the free market solution? If the goal is to have bitcoin be an actual working currency for the people (and it may not be that to you), then it would seem the goal should be to be able to process the vast majority of transactions, instead of just those for the richest half. A "solution" that is economic and solved by free markets tends to be about how to distrubute limited resources. I.e. here's a nice piece of property for sale - who gets it? The highest bidder. problem solved. But I don't think the challenge for bitcoin is how to auction off space in the blockchain, it's how to make that space wide enough to be reliable as a potential currency for common people. If, every time you bought a candy bar at walgreens, you had to pay for the candy bar and take a guess at an extra fee for your payment and hope it goes through 10 minutes later... it just would not work. So the market would be solving something, but it wouldn't be towards the goal of bitcoin being a currency. And honestly I think that's what the opponents of the block size increase want. As cryptocurrency tech evolves, we're seeing that various parties have new self interests to promote. Competing technologies, etc. And they benefit by bitcoin becoming artificially limited. I know there are other reasons, but it's a big part of why this is even a debate.
  - Animats 11y ago
    "What do you mean by this? If 2mb of transaction data is generated every 10 minutes, but only 1mb of it is mined, what is the free market solution?" When you submit a transaction, you're entering an auction. The winners of the auction get their transactions processed quickly. The losers don't. The market thus sets a price on getting a transaction processed. This discourages free-riders who expect their transactions to be processed free or for a nominal fee. Miners are facing the dreaded "halving" next year, when the block reward drops by half. They need higher fees to compensate.[1] [1] http://www.coindesk.com/new-study-low-bitcoin-transaction-fees-unsustainable/ http://www.coindesk.com/new-study-low-bitcoin-transaction-fe...
    - beaner 11y ago
      This isn't an actual solution though, because space is still limited. Bitcoin cannot be a currency if it can only fit 2,000 transactions per block. Even if everyone is willing to pay $100 in fees for their $1 transactions, they won't all fit, and that's broken. The challenge is to find a way for miners to be incentivized and for them also to process all reasonable transactions. There's no reason miner's couldn't increase their fees independently of the block size increase.
    - eyci 11y ago
      > They need higher fees to compensate. That's not an argument for any block size in particular, though. Saying that fee revenue is maximized at smaller blocks is positing that demand for block space (i.e. transaction demand) is relatively inelastic, whereas there isn't really any consensus or even much discussion on what the elasticity is.
    - sneak 11y ago
      Auctions for limited capacity at this ridiculously early stage in bitcoin's growth will shake confidence in the network to the point of abandonment. A currency you can't easily transact in is one nobody wants to use. An auction process like this means you can't easily use bitcoin.
  - AndrewDMcG 11y ago
- reddytowns 11y ago
  I don't think the miners have as much say as you think. If the users all get together and hard fork the protocol, either the miners go along with it, or their blocks will be rejected, no matter how much hashing power they have. The only thing they could reasonably do is go rogue and try to 51% attack the new fork. If they did that successfully, however, they'd be showing that bitcoin isn't as strong as people think it is, thereby destroying their own business (it really is practically being mined by only 5 to 10 miners nowadays). This is why I like alternative coins like Peercoin and Neucoin which are trying to iron out the details of proof-of-stake. Even Ethereum is supposedly going to move to a proof-of-stake model.
  - ars 11y ago
    The miners are the ones creating blocks. Users (i.e. transaction senders) simply have to accept it. They have no input into making or verifying blocks.
    - jnbiche 11y ago
      This is not true. People who run full nodes (as a group) have just as much power as miners. If they all run a hard-forked client, then the miners will be happily mining away on their (worthless) chain.
      - ars 11y ago
        That just isn't true. The miners are the one that decide which chain is "real" and which isn't. The frontend has code to verify the decision of the miner, and it has no ability to change anything.
        reddytowns 11y ago
        >The frontend has code to verify the decision of the miner, and it has no ability to change anything. But, if all the users agree to use frontend code that rejects the decision of the miners that refuse to play along, then the chain the miners are working on is effectively dead.
- mrb 11y ago
  "Unless > 50% of the miners, by hash capacity, make a change that affects the blockchain, it doesn't happen." This is not true at all. If the top 4 pools start accepting and mining 20MB blocks, no one else in the network will accept the blocks, even if that chain is longer, because it would violate the fundamental rule, hard coded in their Bitcoin implementation, that blocks must be < 1MB. It would effectively fork the block chain.
- mirimir 11y ago
  This is off-topic, but bitlaunder.com is arguably a poor choice for Bitcoin mixing, given that it does business in the US. Far better is Bitcoin Fog at foggeddriztrcar2.onion, because coercing them will be harder.
  - jackgavigan 11y ago
    What do you mean by "does business in the US"? Per their website: HEAD OFFICE LOCATION: Moscow Citydel 9 Zemlyanoyval, 4th Floor Moscow 105064 Russian Federation +74959679797
    - mirimir 11y ago
      Sorry, my mistake. Perhaps I was just spreading gossip. But do see http://bitcoin.stackexchange.com/questions/25722/bitlaunders-mixer-vs-major-exchanges-mixer http://bitcoin.stackexchange.com/questions/25722/bitlaunders...
vessenes 11y ago
I have a fairly reductionist view on this, one I think is helpful, so here goes. First, I think of almost all large-scale Bitcoin arguments as religious. Especially if you're an outsider to Bitcoin, it will probably help you to think of them that way, too. People come to a belief, and then back-fill plausible arguments all the time. When there are vigorous debates like this in Bitcoin land, we usually have some strongly held underlying beliefs coming to conflict. I would say the holy war here revolves around a single axiom on each side: the "Small Block" camp fundamentally believes that there is a sort of inflation that happens when block sizes grow. Just like they believed that increased adoption would increase Bitcoin prices, they believe at some level that increased scarcity for block slots will increase transaction fees, and that the system relies on this promise to maintain its integrity. The "Large Block" camp fundamentally believes that the value in the protocol comes from cheap, large-scale transaction support, and that artificially limiting right now is harmful. You can reduce most of the arguments to comments on this. In particular, I think the comments about centralized mining concerns are almost totally specious. Bitcoin has long ago left behind small hobbyist miners with tiny data connections. Block size changes would not have helped them keep up -- there are fundamental aggregation economics in mining that have pushed them out. Any small miner that wants to mine now will do exactly the same thing they would do with 1 Gigabyte blocks: get their work from a pool, and spot verify they haven't been cheated. Embedded in the world view of most "Large Block" miners if you follow out the logic is the belief that probably transaction fees won't pay for the entire security of the network. Ed Felten put out some calculations to this effect while he was at Princeton. They're reasonably compelling arguments, or at least worth thinking hard about. If you give up on the idea that tx fees have to pay for the entire security of the network, I think large blocks become a super obvious choice to enhance the current utility of the network. If you do not want to give up on that idea, then any block size increase decreases the chance that the tx fee scheme will work. Since it is currently not working (the last two blocks as I write this had .09 and .02 BTC in transaction fees respectively), all such changes must be vigorously fought. At any rate, this is why the "small block" folks have a 'purity' vibe when they argue; I think they feel they are protecting the original scheme, one they bought into and care about. As you might be able to tell, I don't think the tx fee scheme is working right now and hence am a "large block" kind of guy. Forcing it to work, (e.g. shrinking block sizes dramatically, and perhaps refusing to mine on blocks with insufficient fees) would dramatically lower the usefulness of Bitcoin to many people. As it is, it's super annoying to have to wonder if a transaction will get included in the next block. Finally, there are some possible tx fee solutions orthogonal to block size miners could implement easily. I have thought for a number of years that miners should just have 'express, regular, bulk' time/fee cutoffs for txs they would include. They should publish these rates. Of course, this could only be done by larger pools, or a consortium of smaller ones. Even 20% would make an impact on the fee schedule. When I think about the economics though, I ask myself 'who would it help?' Miners are already dealing with very difficult economics at a macro and micro level, especially if they have chip projects. Getting a little more juice out of their mining would be nice, but is absolutely not what the large-scale mining game is about. On the flip side, it adds cognitive load and frustration to people transacting. It doesn't make a ton of sense to me. If we're in a world where Bitcoin is still struggling to be useful to a billion people daily, then cognitive load and frustration should be ruthlessly excised from the technology until we get to critical mass.
AndrewDMcG 11y ago
Does the author have any evidence for his claim "The limit was temporary"? It was certainly commonly claimed even a couple of years ago that the coming scarcity of blockchain space would lead to an increase in fees.
- ikeboy 11y ago
  https://bitcointalk.org/index.php?topic=287.msg8810#msg8810 https://bitcointalk.org/index.php?topic=287.msg8810#msg8810 >The eventual solution will be to not care how big it gets. https://bitcointalk.org/index.php?topic=1347.msg15366#msg15366 https://bitcointalk.org/index.php?topic=1347.msg15366#msg153... https://bitcointalk.org/index.php?topic=946236.msg10388435#msg10388435 https://bitcointalk.org/index.php?topic=946236.msg10388435#m... >I'm the guy who went over the blockchain stuff in Satoshi's first cut of the bitcoin code. Satoshi didn't have a 1MB limit in it. The limit was originally Hal Finney's idea. Both Satoshi and I objected that it wouldn't scale at 1MB. Hal was concerned about a potential DoS attack though, and after discussion, Satoshi agreed. The 1MB limit was there by the time Bitcoin launched. But all 3 of us agreed that 1MB had to be temporary because it would never scale.
  - AndrewDMcG 11y ago
    Ah, that's really significant, thanks

46 points by notsony 11y ago

In my view, it's a great and simple short term fix, waiting for more volumes and bigger bucks to get involved. I like it.

itistoday2 11y ago

Relevant: coverage of various perspectives and positions by Ethan Heilman along with counterarguments to each: https://github.com/EthanHeilman/BlockSizeDebate/blob/master/README.md https://github.com/EthanHeilman/BlockSizeDebate/blob/master/... I like that it also includes discussion of the Lightning Network's role, and the idea of making the block size dynamic. These are less heard of perspectives that nevertheless seem significant.

Animats 11y ago

The only people who have a vote on this are the big miners. Unless > 50% of the miners, by hash capacity, make a change that affects the blockchain, it doesn't happen. The top 4 mining operations control more than 50% of the hash rate. Nobody else matters.[1] If the blockchain runs out of capacity, the transactions with the biggest fee attached get processed first. So the free market will solve any overload problem. Bitcoin volume in dollars isn't increasing.[2] The peak was over a year ago. The number of transactions is up, but not the total dollar value. Some of the transactions are just "mixing" transactions, to assist in money laundering. The companies that offer such services are quite open about this.[3] [1] https://blockchain.info/pools https://blockchain.info/pools [2] https://blockchain.info/charts/estimated-transaction-volume-usd https://blockchain.info/charts/estimated-transaction-volume-... [3] https://bitlaunder.com/ https://bitlaunder.com/

beaner 11y ago
> So the free market will solve any overload problem. What do you mean by this? If 2mb of transaction data is generated every 10 minutes, but only 1mb of it is mined, what is the free market solution? If the goal is to have bitcoin be an actual working currency for the people (and it may not be that to you), then it would seem the goal should be to be able to process the vast majority of transactions, instead of just those for the richest half. A "solution" that is economic and solved by free markets tends to be about how to distrubute limited resources. I.e. here's a nice piece of property for sale - who gets it? The highest bidder. problem solved. But I don't think the challenge for bitcoin is how to auction off space in the blockchain, it's how to make that space wide enough to be reliable as a potential currency for common people. If, every time you bought a candy bar at walgreens, you had to pay for the candy bar and take a guess at an extra fee for your payment and hope it goes through 10 minutes later... it just would not work. So the market would be solving something, but it wouldn't be towards the goal of bitcoin being a currency. And honestly I think that's what the opponents of the block size increase want. As cryptocurrency tech evolves, we're seeing that various parties have new self interests to promote. Competing technologies, etc. And they benefit by bitcoin becoming artificially limited. I know there are other reasons, but it's a big part of why this is even a debate.
- Animats 11y ago
  "What do you mean by this? If 2mb of transaction data is generated every 10 minutes, but only 1mb of it is mined, what is the free market solution?" When you submit a transaction, you're entering an auction. The winners of the auction get their transactions processed quickly. The losers don't. The market thus sets a price on getting a transaction processed. This discourages free-riders who expect their transactions to be processed free or for a nominal fee. Miners are facing the dreaded "halving" next year, when the block reward drops by half. They need higher fees to compensate.[1] [1] http://www.coindesk.com/new-study-low-bitcoin-transaction-fees-unsustainable/ http://www.coindesk.com/new-study-low-bitcoin-transaction-fe...
  - beaner 11y ago
    This isn't an actual solution though, because space is still limited. Bitcoin cannot be a currency if it can only fit 2,000 transactions per block. Even if everyone is willing to pay $100 in fees for their $1 transactions, they won't all fit, and that's broken. The challenge is to find a way for miners to be incentivized and for them also to process all reasonable transactions. There's no reason miner's couldn't increase their fees independently of the block size increase.
  - eyci 11y ago
    > They need higher fees to compensate. That's not an argument for any block size in particular, though. Saying that fee revenue is maximized at smaller blocks is positing that demand for block space (i.e. transaction demand) is relatively inelastic, whereas there isn't really any consensus or even much discussion on what the elasticity is.
  - sneak 11y ago
    Auctions for limited capacity at this ridiculously early stage in bitcoin's growth will shake confidence in the network to the point of abandonment. A currency you can't easily transact in is one nobody wants to use. An auction process like this means you can't easily use bitcoin.
- AndrewDMcG 11y ago
reddytowns 11y ago
I don't think the miners have as much say as you think. If the users all get together and hard fork the protocol, either the miners go along with it, or their blocks will be rejected, no matter how much hashing power they have. The only thing they could reasonably do is go rogue and try to 51% attack the new fork. If they did that successfully, however, they'd be showing that bitcoin isn't as strong as people think it is, thereby destroying their own business (it really is practically being mined by only 5 to 10 miners nowadays). This is why I like alternative coins like Peercoin and Neucoin which are trying to iron out the details of proof-of-stake. Even Ethereum is supposedly going to move to a proof-of-stake model.
- ars 11y ago
  The miners are the ones creating blocks. Users (i.e. transaction senders) simply have to accept it. They have no input into making or verifying blocks.
  - jnbiche 11y ago
    This is not true. People who run full nodes (as a group) have just as much power as miners. If they all run a hard-forked client, then the miners will be happily mining away on their (worthless) chain.
    - ars 11y ago
      That just isn't true. The miners are the one that decide which chain is "real" and which isn't. The frontend has code to verify the decision of the miner, and it has no ability to change anything.
      - reddytowns 11y ago
        >The frontend has code to verify the decision of the miner, and it has no ability to change anything. But, if all the users agree to use frontend code that rejects the decision of the miners that refuse to play along, then the chain the miners are working on is effectively dead.
mrb 11y ago
"Unless > 50% of the miners, by hash capacity, make a change that affects the blockchain, it doesn't happen." This is not true at all. If the top 4 pools start accepting and mining 20MB blocks, no one else in the network will accept the blocks, even if that chain is longer, because it would violate the fundamental rule, hard coded in their Bitcoin implementation, that blocks must be < 1MB. It would effectively fork the block chain.
mirimir 11y ago
This is off-topic, but bitlaunder.com is arguably a poor choice for Bitcoin mixing, given that it does business in the US. Far better is Bitcoin Fog at foggeddriztrcar2.onion, because coercing them will be harder.
- jackgavigan 11y ago
  What do you mean by "does business in the US"? Per their website: HEAD OFFICE LOCATION: Moscow Citydel 9 Zemlyanoyval, 4th Floor Moscow 105064 Russian Federation +74959679797
  - mirimir 11y ago
    Sorry, my mistake. Perhaps I was just spreading gossip. But do see http://bitcoin.stackexchange.com/questions/25722/bitlaunders-mixer-vs-major-exchanges-mixer http://bitcoin.stackexchange.com/questions/25722/bitlaunders...

vessenes 11y ago

I have a fairly reductionist view on this, one I think is helpful, so here goes. First, I think of almost all large-scale Bitcoin arguments as religious. Especially if you're an outsider to Bitcoin, it will probably help you to think of them that way, too. People come to a belief, and then back-fill plausible arguments all the time. When there are vigorous debates like this in Bitcoin land, we usually have some strongly held underlying beliefs coming to conflict. I would say the holy war here revolves around a single axiom on each side: the "Small Block" camp fundamentally believes that there is a sort of inflation that happens when block sizes grow. Just like they believed that increased adoption would increase Bitcoin prices, they believe at some level that increased scarcity for block slots will increase transaction fees, and that the system relies on this promise to maintain its integrity. The "Large Block" camp fundamentally believes that the value in the protocol comes from cheap, large-scale transaction support, and that artificially limiting right now is harmful. You can reduce most of the arguments to comments on this. In particular, I think the comments about centralized mining concerns are almost totally specious. Bitcoin has long ago left behind small hobbyist miners with tiny data connections. Block size changes would not have helped them keep up -- there are fundamental aggregation economics in mining that have pushed them out. Any small miner that wants to mine now will do exactly the same thing they would do with 1 Gigabyte blocks: get their work from a pool, and spot verify they haven't been cheated. Embedded in the world view of most "Large Block" miners if you follow out the logic is the belief that probably transaction fees won't pay for the entire security of the network. Ed Felten put out some calculations to this effect while he was at Princeton. They're reasonably compelling arguments, or at least worth thinking hard about. If you give up on the idea that tx fees have to pay for the entire security of the network, I think large blocks become a super obvious choice to enhance the current utility of the network. If you do not want to give up on that idea, then any block size increase decreases the chance that the tx fee scheme will work. Since it is currently not working (the last two blocks as I write this had .09 and .02 BTC in transaction fees respectively), all such changes must be vigorously fought. At any rate, this is why the "small block" folks have a 'purity' vibe when they argue; I think they feel they are protecting the original scheme, one they bought into and care about. As you might be able to tell, I don't think the tx fee scheme is working right now and hence am a "large block" kind of guy. Forcing it to work, (e.g. shrinking block sizes dramatically, and perhaps refusing to mine on blocks with insufficient fees) would dramatically lower the usefulness of Bitcoin to many people. As it is, it's super annoying to have to wonder if a transaction will get included in the next block. Finally, there are some possible tx fee solutions orthogonal to block size miners could implement easily. I have thought for a number of years that miners should just have 'express, regular, bulk' time/fee cutoffs for txs they would include. They should publish these rates. Of course, this could only be done by larger pools, or a consortium of smaller ones. Even 20% would make an impact on the fee schedule. When I think about the economics though, I ask myself 'who would it help?' Miners are already dealing with very difficult economics at a macro and micro level, especially if they have chip projects. Getting a little more juice out of their mining would be nice, but is absolutely not what the large-scale mining game is about. On the flip side, it adds cognitive load and frustration to people transacting. It doesn't make a ton of sense to me. If we're in a world where Bitcoin is still struggling to be useful to a billion people daily, then cognitive load and frustration should be ruthlessly excised from the technology until we get to critical mass.

AndrewDMcG 11y ago

Does the author have any evidence for his claim "The limit was temporary"? It was certainly commonly claimed even a couple of years ago that the coming scarcity of blockchain space would lead to an increase in fees.

ikeboy 11y ago
https://bitcointalk.org/index.php?topic=287.msg8810#msg8810 https://bitcointalk.org/index.php?topic=287.msg8810#msg8810 >The eventual solution will be to not care how big it gets. https://bitcointalk.org/index.php?topic=1347.msg15366#msg15366 https://bitcointalk.org/index.php?topic=1347.msg15366#msg153... https://bitcointalk.org/index.php?topic=946236.msg10388435#msg10388435 https://bitcointalk.org/index.php?topic=946236.msg10388435#m... >I'm the guy who went over the blockchain stuff in Satoshi's first cut of the bitcoin code. Satoshi didn't have a 1MB limit in it. The limit was originally Hal Finney's idea. Both Satoshi and I objected that it wouldn't scale at 1MB. Hal was concerned about a potential DoS attack though, and after discussion, Satoshi agreed. The 1MB limit was there by the time Bitcoin launched. But all 3 of us agreed that 1MB had to be temporary because it would never scale.
- AndrewDMcG 11y ago
  Ah, that's really significant, thanks