|
Title: Why did this transaction confirm? Post by: Satosh¡ Slot on July 08, 2014, 11:18:00 AM https://blockchain.info/tx/ee61b911610f66f539832699fbbf4ab3955c8bd5ad0cfa570ff500dedcde5bf8
Isn't the 1 satoshi outputs present in this transaction supposed to be considered dust? The transaction fee is only 0.0001 for a 26kb transaction, it appears very low so I assume this transaction had very low priority. What am I missing? Title: Re: Why did this transaction confirm? Post by: amaclin on July 08, 2014, 12:29:04 PM Quote Re: Why did this transaction confirm? 1) because it is valid 2) because one of miners decided to include it to a block Quote Isn't the 1 satoshi outputs present in this transaction supposed to be considered dust? Dust outputs are allowed. Standard clients disallows to send dust values without fees, but this transaction was not created by standard client, but some kind of bot-software Quote The transaction fee is only 0.0001 for a 26kb transaction, it appears very low so I assume this transaction had very low priority. Yes. But if you are miner you are able to put in block whatever you want while it is valid. Another way - is to find a miner with old (but still not obsolete) software Title: Re: Why did this transaction confirm? Post by: gmaxwell on July 08, 2014, 12:51:26 PM I've already nagged the miner in question about the wad of dust floods they just mined.
If you've received some of this 1e-8 dust to a bitcoin-core wallet, https://github.com/petertodd/dust-b-gone is a good way to get rid of it. Title: Re: Why did this transaction confirm? Post by: Satosh¡ Slot on July 08, 2014, 01:47:31 PM Quote Re: Why did this transaction confirm? 1) because it is valid 2) because one of miners decided to include it to a block Yes, but I saw earlier transactions take a long time to be included in a block and blockchain lists this as received and included in a block at the same time. So I guess it had problems being brodcasted properly, but still got included in a block? Or was it mined the same minute it was created? Title: Re: Why did this transaction confirm? Post by: gmaxwell on July 08, 2014, 01:50:21 PM Yes, but I saw earlier transactions take a long time to be included in a block and blockchain lists this as received and included in a block at the same time. So I guess it had problems being brodcasted properly, but still got included in a block? Or was it mined the same minute it was created? The times reported on bc.i are just whenever they saw it. Because this is a dust transaction that wouldn't have generally been relayed (and because bc.i themselves suppress these when unconfirmed) it wouldn't have been seen by bc.i until it ended up in a block.Title: Re: Why did this transaction confirm? Post by: DeathAndTaxes on July 08, 2014, 02:48:45 PM As pointed out dust is allowed. That miner just massively bloated the UTXO and those outputs will very likely never be spent. The protocol tries to make this difficult by making dust non-standard and it assumes that miners are at least moderately intelligent. Miners have to parse the UTXO to verify txn so including this dusty garbage only makes their future work more difficult. In a logical world with rational miners wouldn't be cutting their own throat.
You may want to ask bcpool.io why they included this txn knowing that nobody will ever use those outputs and thus they will bloat the UTXO (increasing memory pressure for every node both now and in the future) probably for perpetuity. Are they malicious or are they just ignorant and still feel competent enough to run a pool? If you don't like this then really there are only two things you can do: a) if you are a miner, then boycott bcpool.io as they are either idiots of malicious. If they have no hashpower they can't mine any more stupid txns. b) Push to make txn with outputs below the dust limit invalid. This would be a hard fork but it could be done. One option would be to have a lower threshold for valid. i.e. min fee is 10,000 satoshis, dust limit is 5,430 sat, valid limit is 543 sat (1/10th dust limit). Title: Re: Why did this transaction confirm? Post by: amaclin on July 08, 2014, 03:00:23 PM Quote If you don't like this then really there are only two things you can do: a) if you are a miner, then boycott bcpool.io as they are either idiots of malicious. Slow down, man. The miner had a right to mine this transaction. And the whole bitcoin network agreed that block is valid. There are a lot of dust and spam transactions in the blockchain. We should live with them and think about future, not about past Title: Re: Why did this transaction confirm? Post by: DeathAndTaxes on July 08, 2014, 03:15:47 PM There are a lot of dust and spam transactions in the blockchain. We should live with them and think about future, not about past Well the pool operator decided to include the txn not the individual miners. Miners have a right to use another pool. If they agree with the decision to include this txn well they obviously wouldn't change pools would they. However if they don't agree with that stupid decision then they are free to use a pool they find more responsible. Quote There are a lot of dust and spam transactions in the blockchain. We should live with them and think about future, not about past That is the whole point isn't it. Most of the dust in the UTXO (forget the blockchain the critical resource is the UTXO) is from BEFORE the dust limits. Since the dust limits were put in place most miners intentionally DO NOT include these dust outputs because they are unspendable and they will bloat the UTXO and increase the cost of all nodes forever. So it is thinking about the future. Miners who don't want to see the UTXO bloated IN THE FUTURE should probably choose another pool. Right?Title: Re: Why did this transaction confirm? Post by: Peter R on July 08, 2014, 05:42:12 PM Hopefully we can build community awareness around the important difference between "blockchain bloat" and "UTXO bloat"--I think many users erroneously lump these two topics as one in their minds. A more intelligent way to limit unnecessary growth of the UTXO set like we just saw with ee61b911... (https://blockchain.info/tx/ee61b911610f66f539832699fbbf4ab3955c8bd5ad0cfa570ff500dedcde5bf8) might be one of the few hard forks I'd support.
By the way, recent analysis of the UTXO set (https://bitcointalk.org/index.php?topic=563972.msg7658989#msg7658989) by DeathAndTaxes as of Block #305,303 indicates that the "dust rule" has had a notable impact. If dust rules had been in place since 2009, we'd have a much cleaner UTXO set today with essentially no loss of function to bitcoin. https://i.imgur.com/j4XjZgN.gif I've already nagged the miner in question about the wad of dust floods they just mined. Blockchain.info indicates that the block with the offending TX (Block #309,740) was relayed by "Unknown 37.187.90.171" and is the only block relayed by that IP address in the last 24 hours. May I ask how you were able to contact them? Title: Re: Why did this transaction confirm? Post by: gmaxwell on July 08, 2014, 05:47:52 PM Blockchain.info indicates that the block with the offending TX (Block #309,740) was relayed by "Unknown 37.187.90.171" and is the only block relayed by that IP address in the last 24 hours. May I ask how you were able to contact them? BC.i's analysis is ... seldom good. Just look in the coinbase of the block.Title: Re: Why did this transaction confirm? Post by: amaclin on July 08, 2014, 06:36:27 PM By the way.
This is not new spam-attack from enjoy-sochi. It is echo of past events. Transaction https://blockchain.info/tx/ee61b911610f66f539832699fbbf4ab3955c8bd5ad0cfa570ff500dedcde5bf8 was created and sent in February. You can find in in google archives. For example http://webcache.googleusercontent.com/search?q=cache:WhRSCSqOMC4J:hibitcoin.com/bitcoin/unspent/18qQzy6SR6HmMpCLs1nHb2jraTMoXPPkW6 Some node ( may be not bcpool.io ) keeped it in its memory pool and relayed today bcpool.io got it, checked and included into a block. Title: Re: Why did this transaction confirm? Post by: gmaxwell on July 08, 2014, 06:55:36 PM Some node ( may be not bcpool.io ) keeped it in its memory pool and relayed today Normal nodes do not relay old transactions... but, indeed, could have just been them not updated after all this time.bcpool.io got it, checked and included into a block. The one I searched for got not hits, good sleuthing. :) Title: Re: Why did this transaction confirm? Post by: Satosh¡ Slot on July 08, 2014, 07:06:07 PM If you've received some of this 1e-8 dust to a bitcoin-core wallet, https://github.com/petertodd/dust-b-gone is a good way to get rid of it. The only person I know that got one of these say he/she has no other unspent output on this address so I guess it's easier to just remove it from his/hers bitcoin core. Still curious what kind of problems he/she can run into when trying to spend it though. Since the dust limits were put in place most miners intentionally DO NOT include these dust outputs because they are unspendable and they will bloat the UTXO and increase the cost of all nodes forever. I see several of these outputs are already spent. How are they supposed to be unspendable? Title: Re: Why did this transaction confirm? Post by: gmaxwell on July 08, 2014, 07:07:14 PM I see several of these outputs are already spent. How are they supposed to be unspendable? They are technically spendable but they cost more in marginal fees than they are worth... It's altruistic to redeem them but not rational to do so otherwise.Title: Re: Why did this transaction confirm? Post by: amaclin on July 08, 2014, 07:07:47 PM Quote Normal nodes do not relay old transactions... They have a "mempool" request and everyone is able to take all transactions from a node and re-broadcast them to the network.You know that it is very easy to do. Even my buggy-script do it for some transactions sometimes. ;D Title: Re: Why did this transaction confirm? Post by: amaclin on July 08, 2014, 07:11:17 PM Quote They are technically spendable but they cost more in marginal fees than they are worth... No. It is possible to collect dust-outputs with free-transactions.Look here https://blockchain.info/address/155M7TvBRww6WFdtGQgTYUH8DuLheNafCf Title: Re: Why did this transaction confirm? Post by: gmaxwell on July 08, 2014, 07:12:56 PM Quote They are technically spendable but they cost more in marginal fees than they are worth... No. It is possible to collect dust-outputs with free-transactions.Look here https://blockchain.info/address/155M7TvBRww6WFdtGQgTYUH8DuLheNafCf Title: Re: Why did this transaction confirm? Post by: amaclin on July 08, 2014, 07:19:46 PM Quote and the loss of priority will me that person must pay more fees in the future. What are you talking about? This person can send free transactions each second block!309655 https://blockchain.info/tx/0e7f9ab6e5b2916b390f749c6b980f669e1dc1e02fc9d795b833451889524ded 309657 https://blockchain.info/tx/bddbbc8d0e9d8dc8e4fd1b48bfd9582318648ba501243015887082773b48a860 309659 https://blockchain.info/tx/ce9862e3cadc5d9b87da119b52b8f2b1c6a6a7e6d86d9df108febcc715c5a89d S/he does not pay fees at all! Title: Re: Why did this transaction confirm? Post by: Peter R on July 08, 2014, 08:17:05 PM It looks like https://blockchain.info/tx/ee61b911610f66f539832699fbbf4ab3955c8bd5ad0cfa570ff500dedcde5bf8 is just one of many spam transactions that have been mined recently. If you look up the 1Enjoy address, you'll see that several transactions were included in blocks 309740 and 309657. Both these blocks were mined by the same group, as the coinbase was awarded to https://blockchain.info/address/19RZPWsNmVM2cVVDMtkzo2UtkFBFc6nFE8 in each case. BCPool.io (https://bcpool.io/index.php?page=statistics&action=blocks) shows that they were the pool responsible for mining these two blocks. In fact, as far as I can tell, these are the only blocks ever mined by BCPool.
In total, 17,976 satoshi dust outputs sent from 1Enjoy have been mined in the last two days by BCPool.io. Code: BLOCK #309740: 10,486 DUST OUTPUTS Title: Re: Why did this transaction confirm? Post by: DannyHamilton on July 08, 2014, 08:41:29 PM It looks like https://blockchain.info/tx/ee61b911610f66f539832699fbbf4ab3955c8bd5ad0cfa570ff500dedcde5bf8 is just one of many spam transactions that have been mined recently. Yep. I noticed this one last week: https://blockchain.info/tx/cff61bad78a1759cf71cf287d67f3a6d63f9f9c2f0566735c06446455834b231 Title: Re: Why did this transaction confirm? Post by: Peter R on July 08, 2014, 08:56:03 PM It looks like https://blockchain.info/tx/ee61b911610f66f539832699fbbf4ab3955c8bd5ad0cfa570ff500dedcde5bf8 is just one of many spam transactions that have been mined recently. Yep. I noticed this one last week: https://blockchain.info/tx/cff61bad78a1759cf71cf287d67f3a6d63f9f9c2f0566735c06446455834b231 Interesting. That one was sent from "SPAM SEND" (1DkRWKXpfQwm5b1pGL2QdpPdrUdKSdzbnG) and mined by Discuss Fish. I checked the two blocks mined recently by BCPool (blocks 309740 and 309657) and indeed they mined another 120 satoshi spam outputs from "SPAM SEND" in addition to the 17,976 from 1Enjoy that I reported earlier. What's interesting is that BCPool also includes some transactions that consolidated a bunch of dust too: https://blockchain.info/address/1MQQAxLKLKt3RznFMLmxXMieEuCqKXbMoe So, who is BCPool? Title: Re: Why did this transaction confirm? Post by: DeathAndTaxes on July 08, 2014, 08:57:55 PM The coinbase txn decoded contains bcpoool.io
https://bcpool.io/ Title: Re: Why did this transaction confirm? Post by: Peter R on July 08, 2014, 09:21:59 PM The coinbase txn decoded contains bcpoool.io https://bcpool.io/ Yes, but there seems to be no information about bcpool.io. Their "about" page: https://bcpool.io/index.php?page=about&action=pool Is also exactly the same this "about" page: https://chunkypools.com/aur/index.php?page=about&action=pool And doesn't contain any information about them at all. Title: Re: Why did this transaction confirm? Post by: amaclin on July 09, 2014, 03:45:22 AM Hmm. I also see some strange things that have never seen before.
Lets take transaction https://blockchain.info/tx/4af1a19fd6940758e2c4a3e779cb25023d93668c75a313f93e8008c0b075488e it was created ant sent a ~week ago with 0 fees and very-very low priority (date on bc.i is incorrect - tx is older) No miner wanted to include this transaction into a block. There was no reason to include this transaction Yesterday somebody re-broadcasted it (bc.i forgot the first copy) again Yesterday I redeemed it with https://blockchain.info/tx/8a1f0f07fb397b2aa0e77995723f2dac9aee7df63a6f05a7db7d168d62a60f94 with a fee of 0.0001 So, there is a reason now for a miner to include both transactions - miner gains 0.0001 (which is more than default 0.00001 per kb) And these two transactions were mined! Of course, in same block (309837) because miner would not include first transaction without second. ---- Update 1: No. Block 309837 contains a lot of free transactions. ---- Update 2: child-pays-for-parent? Title: Re: Why did this transaction confirm? Post by: RedDiamond on July 09, 2014, 02:38:25 PM Hmm. I also see some strange things that have never seen before. Lets take transaction https://blockchain.info/tx/4af1a19fd6940758e2c4a3e779cb25023d93668c75a313f93e8008c0b075488e Some of the addresses in transaction are also quite interesting: 1P1ease..., 1Give..., 1hint... Title: Re: Why did this transaction confirm? Post by: amaclin on July 09, 2014, 03:06:30 PM Quote Some of the addresses in transaction are also quite interesting: 1P1ease..., 1Give..., 1hint... Nothing interesting. You have missed topic https://bitcointalk.org/index.php?topic=661781.0 (I do not remember the exact page) The transaction with vanity addresses was used as a "public message". Title: Re: Why did this transaction confirm? Post by: RedDiamond on July 09, 2014, 03:42:37 PM Quote Some of the addresses in transaction are also quite interesting: 1P1ease..., 1Give..., 1hint... Nothing interesting. You have missed topic https://bitcointalk.org/index.php?topic=661781.0 (I do not remember the exact page) The transaction with vanity addresses was used as a "public message". Yes, missed totally :-[. Thank you for the link, looks that the puzzle was great fun. Title: Re: Why did this transaction confirm? Post by: runeks on July 26, 2014, 02:02:40 PM FWIW, one of the outputs in f75b455b4df94bcdcd54985b4aeea9948d2a1dca0f63c65b85ad2d941050c5cf sends 1 Satoshi to an address I own, that has been used as a change address in a previous transaction. So it looks like the outputs aren't random, or at least not all of them. Someone has scanned the blockchain are they are sending outputs to existing addresses.
I discovered this thread because I noticed someone had sent 1 Satoshi to me on July 8, and the transaction ID led me to this thread. |
