Speculate as to how this will play out
http://www.ibtimes.co.uk/coinwallet-plans-bitcoin-dust-attack-september-create-30-day-transaction-backlog-1515981
Quote
UK-based mining service CoinWallet is gearing up to conduct a stress test of the Bitcoin network in early September, which it said will likely render most standard wallet software "worthless" and create "nearly a 30-day backlog".
A CoinWallet representative told IBTimes in an email exchange: "I don't have a set date, but it will be early September. I'm too busy this month to fully devote a large amount of time to executing the 'test'.
"Despite the general belief that someone can do something like this by setting up a couple of servers to send payments to yourself over and over, it's a little more complicated than that."
CoinWallet's last test of the system involved taking a bunch of Bitcoin and then dividing it among thousands of Bitcoin addresses, all in tiny amounts of 0.00001.
"As part of this test, I will be reconsolidating more than 150 Bitcoin that currently sits in these wallets."
CoinWallet conducted a transaction to demonstrate what this will look like.
"As you can see from the transaction, there are 20 tiny inputs, with half going to miner fees, and half going to one of my CoinWallet addresses. This transaction is approximately 3kb, or 1/323rd of a block. In other words, for every ~323 of these I send, I fill up a block.
"Each of these transactions is for a total amount of 0.0002 BTC (including miner fees), meaning that with my 150 btc that is currently in this state, I will be crafting 750000 transactions, or 2321 blocks, or a 16 day backlog."
CoinWallet said the second phase of the test will be to use 20 cloud-hosted servers to send bitcoin payments in amounts of 0.00001 to thousands of its addresses at random, all in transactions that are also approximately 3kb each.
"These 20 servers push approximately 1 transaction per second. The plan is to fill them up to 50-100 Bitcoin in total. In theory, if all things go as planned, we will create a nearly 30-day backlog."
"Of course, this won't cripple Bitcoin entirely. Those who are smart enough to increase their fees will still manage to push transactions through. However, it will make it prohibitively expensive, and will likely render most standard wallet software, ranging from Multibit, to Mycellium, Blockchain.info and others completely worthless.
"I should however note that CoinWallet.eu is relatively immune to this form of 'attack' as our fees are dynamic and are set at 3x the standard limit. As always, CoinWallet clients will be unaffected by the test. More details will be posted publicly when the test is imminent."
The WSJ BitBeat reported in July that CoinWallet had communicated plans to flood the system with countless small transactions, enough to fill two blocks every minute, and after that send out slightly larger transactions. Its goal was to create a 200 megabyte blockchain backlog.
The previous CoinWallet test only achieved about 15% of its output because its servers crashed.
"We feel that our tests might prove to be the catalyst that propels the core devs and miners to implement the required hard fork that is desperately needed," Coinwallet told BitBeat.
A debate has been raging for months over whether or not to increase the maximum size of a transaction block of data beyond 1 megabyte.
Bitcoin core developers Gavin Andresen and Mike Hearn are spearheading the drive to increase the block size, and have developed Bitcoin XT client to allow miners to opt out of the current 1MB limit. The majority of miners must adopt the protocol upgrade or else no change will come into effect.
"The fact that the XT fork hasn't occurred yet is ridiculous," CoinWallet said.
A CoinWallet representative told IBTimes in an email exchange: "I don't have a set date, but it will be early September. I'm too busy this month to fully devote a large amount of time to executing the 'test'.
"Despite the general belief that someone can do something like this by setting up a couple of servers to send payments to yourself over and over, it's a little more complicated than that."
CoinWallet's last test of the system involved taking a bunch of Bitcoin and then dividing it among thousands of Bitcoin addresses, all in tiny amounts of 0.00001.
"As part of this test, I will be reconsolidating more than 150 Bitcoin that currently sits in these wallets."
CoinWallet conducted a transaction to demonstrate what this will look like.
"As you can see from the transaction, there are 20 tiny inputs, with half going to miner fees, and half going to one of my CoinWallet addresses. This transaction is approximately 3kb, or 1/323rd of a block. In other words, for every ~323 of these I send, I fill up a block.
"Each of these transactions is for a total amount of 0.0002 BTC (including miner fees), meaning that with my 150 btc that is currently in this state, I will be crafting 750000 transactions, or 2321 blocks, or a 16 day backlog."
CoinWallet said the second phase of the test will be to use 20 cloud-hosted servers to send bitcoin payments in amounts of 0.00001 to thousands of its addresses at random, all in transactions that are also approximately 3kb each.
"These 20 servers push approximately 1 transaction per second. The plan is to fill them up to 50-100 Bitcoin in total. In theory, if all things go as planned, we will create a nearly 30-day backlog."
"Of course, this won't cripple Bitcoin entirely. Those who are smart enough to increase their fees will still manage to push transactions through. However, it will make it prohibitively expensive, and will likely render most standard wallet software, ranging from Multibit, to Mycellium, Blockchain.info and others completely worthless.
"I should however note that CoinWallet.eu is relatively immune to this form of 'attack' as our fees are dynamic and are set at 3x the standard limit. As always, CoinWallet clients will be unaffected by the test. More details will be posted publicly when the test is imminent."
The WSJ BitBeat reported in July that CoinWallet had communicated plans to flood the system with countless small transactions, enough to fill two blocks every minute, and after that send out slightly larger transactions. Its goal was to create a 200 megabyte blockchain backlog.
The previous CoinWallet test only achieved about 15% of its output because its servers crashed.
"We feel that our tests might prove to be the catalyst that propels the core devs and miners to implement the required hard fork that is desperately needed," Coinwallet told BitBeat.
A debate has been raging for months over whether or not to increase the maximum size of a transaction block of data beyond 1 megabyte.
Bitcoin core developers Gavin Andresen and Mike Hearn are spearheading the drive to increase the block size, and have developed Bitcoin XT client to allow miners to opt out of the current 1MB limit. The majority of miners must adopt the protocol upgrade or else no change will come into effect.
"The fact that the XT fork hasn't occurred yet is ridiculous," CoinWallet said.