3 ms·

Pretty good article, and a great overview of the blocksize controversy. Also a great exhibit of some stereotypical programmer social problems; we don't get a l

Pretty good article, and a great overview of the blocksize controversy. Also a great exhibit of some stereotypical programmer social problems; we don't get a lot of middle ground, most programmers are either openly hostile and combative or so deathly afraid of confrontation and responsibility that they give away their authority so that they don't have to handle the pressure. Gavin should've kept control. Bitcoin is learning exactly why a strong central authority is so desirable in money exchange: it keeps the value of the currency stable by preventing panic and confusion over issues like this. Many discussions of Bitcoin claim that this power is transferred to "the network" to make the final decision, which sounds very egalitarian and democratic, but Bitcoin failed to provide the controls that would prevent power hoarding and ensure that the people who depend on bitcoin were fairly represented. This is one reason why modern democracies are structured within the framework of a republic. This is probably one of bitcon's hardest to solve problems, since the hardware to get respectable hashrates is unobtainable for quite literally everyone who doesn't have access to their own microfabrication facilities. Even if one of the specialty bitcoin hardware makers had a really good, cheap chip, why would they share it? They'd hoard all the hashpower for themselves. Litecoin attempts to address this by hasing with scrypt, under the belief that it's harder to hoard power with custom hardware if the algorithm uses a lot of processing power and memory instead of just a lot of processing power. Mike failed to mention one incentive that exists to prevent increasing the block size: miners get the transaction fees attached to each block they mine. If the block size is large, there is little contention for space in the blocks, and ergo there is not much reason to incentivize miners to include your transaction in the next block. By keeping the artificial constraint on the block size, people who own a lot of hash power will be gaining a lot more bitcoin for themselves. I don't think this crisis is insurmountable. So much money has been sunk into bitcoin that I can't believe people are just going to let this cabal take it out. BitcoinXT will gain notoriety through the mainstream press and the resultant sell off among casual investors will freak the big players out and force them into running XT nodes.

MBCook 11y ago
> Mike failed to mention one incentive that exists to prevent increasing the block size: miners get the transaction fees attached to each block they mine. If the block size is large, there is little contention for space in the blocks, and ergo there is not much reason to incentivize miners to include your transaction in the next block. By keeping the artificial constraint on the block size, people who own a lot of hash power will be gaining a lot more bitcoin If the blocks are bigger, doesn't that mean they can hold more transactions and thus get more fees total? Or are you assuming the relative scarcity of transaction space would push fees high enough to overcome that?
- cookiecaper 11y ago
  >are you assuming the relative scarcity of transaction space would push fees high enough to overcome that? Yes. The blocks always contain approximately 10 minutes' worth of transactions because the network periodically adjusts mining difficulty to approximate that target. Right now, the issue is that there are times when 10 minutes' worth of bitcoin transactions are occupying more than 1MB of space in the completed blocks, which means a processing backlog is formed. Miners prioritize transactions by their attached transaction fee, since they get that fee if they find the block. Thus, users are effectively placing a bid for the network hash power to verify their transaction. When a backlog forms, customers that want their transaction processed quickly have to outbid others to get a miner to start working on it. If there's not a backlog, their transaction will be included with only a token fee attached because something is better than nothing. Going from 1MB to 8MB means that bitcoin would need 8x more transaction volume per 10 minute block to get back into a transaction backlog, which is the only time that users will attach a meaningful transaction fee to their transactions. So if the block size goes up from 1MB to 8MB today, there will still be a block every ten minutes, but since transaction volume will presumably remain nearly the same, space in each block won't be scarce and the bidding war won't take off. It'd be a long time before we got back into the same predicament, meaning miners would have to wait a lot longer to start collecting meaningful transaction fees. That's why some people with heavy investments in mining want to keep the block size artificially low: they're trying to instigate a bidding war for their hash power. It should be noted that this eventuality was always part of bitcoin's design. Bitcoin is programmed to stop "minting" around 22 million coins. At that time, the network will not issue any reward to the miners that find a block (the reward will cut itself down until that number is reached, targeted for approximately 2022 iirc). The solution to this has always been "users will have to incentivize miners with transaction fees". It's just that the assumption was always that we wouldn't have to deal with that until the network itself stopped attaching rewards to mining. In practice, however, we're in that situation now due to the artificial constraint on the size of a block (which, afaik, is mostly accidental and was never intended to be permanent). The debate is over whether bitcoin should remove the artificial block size constraint and keep its fees negligibly low or whether it should keep the constraint and "allow" miners to start charging more for processing transactions.

MBCook 11y ago

> Mike failed to mention one incentive that exists to prevent increasing the block size: miners get the transaction fees attached to each block they mine. If the block size is large, there is little contention for space in the blocks, and ergo there is not much reason to incentivize miners to include your transaction in the next block. By keeping the artificial constraint on the block size, people who own a lot of hash power will be gaining a lot more bitcoin If the blocks are bigger, doesn't that mean they can hold more transactions and thus get more fees total? Or are you assuming the relative scarcity of transaction space would push fees high enough to overcome that?

cookiecaper 11y ago
>are you assuming the relative scarcity of transaction space would push fees high enough to overcome that? Yes. The blocks always contain approximately 10 minutes' worth of transactions because the network periodically adjusts mining difficulty to approximate that target. Right now, the issue is that there are times when 10 minutes' worth of bitcoin transactions are occupying more than 1MB of space in the completed blocks, which means a processing backlog is formed. Miners prioritize transactions by their attached transaction fee, since they get that fee if they find the block. Thus, users are effectively placing a bid for the network hash power to verify their transaction. When a backlog forms, customers that want their transaction processed quickly have to outbid others to get a miner to start working on it. If there's not a backlog, their transaction will be included with only a token fee attached because something is better than nothing. Going from 1MB to 8MB means that bitcoin would need 8x more transaction volume per 10 minute block to get back into a transaction backlog, which is the only time that users will attach a meaningful transaction fee to their transactions. So if the block size goes up from 1MB to 8MB today, there will still be a block every ten minutes, but since transaction volume will presumably remain nearly the same, space in each block won't be scarce and the bidding war won't take off. It'd be a long time before we got back into the same predicament, meaning miners would have to wait a lot longer to start collecting meaningful transaction fees. That's why some people with heavy investments in mining want to keep the block size artificially low: they're trying to instigate a bidding war for their hash power. It should be noted that this eventuality was always part of bitcoin's design. Bitcoin is programmed to stop "minting" around 22 million coins. At that time, the network will not issue any reward to the miners that find a block (the reward will cut itself down until that number is reached, targeted for approximately 2022 iirc). The solution to this has always been "users will have to incentivize miners with transaction fees". It's just that the assumption was always that we wouldn't have to deal with that until the network itself stopped attaching rewards to mining. In practice, however, we're in that situation now due to the artificial constraint on the size of a block (which, afaik, is mostly accidental and was never intended to be permanent). The debate is over whether bitcoin should remove the artificial block size constraint and keep its fees negligibly low or whether it should keep the constraint and "allow" miners to start charging more for processing transactions.