must have been a glitch in the system
That's a useless statement, basically.
I'm not sure if it's related, but the Chinese pools are the ones pushing back against a block size increase due to limited bandwidth.
Topic: Blocks 364629, 364632, 364633 and 364634 are all empty (Read 821 times)
That's a useless statement, basically.
I'm not sure if it's related, but the Chinese pools are the ones pushing back against a block size increase due to limited bandwidth.
must have been a glitch in the system
...
Thank you for the education guys.
Since I am not a programmer nor have any friends or family who are into BTC, I have to observe and read (bitcointalk) the best I can.
And ask questions/make observations about things I do not understand (or dimly so).
Thank you for the education guys.
Since I am not a programmer nor have any friends or family who are into BTC, I have to observe and read (bitcointalk) the best I can.
And ask questions/make observations about things I do not understand (or dimly so).
From my understanding, here is the basic "proper" operation of a miner
The reason why step (2) exists is that step (4) can take MINUTES.... Now we all hope that the machine running getblocktemplate is the biggest fastest meanest manchine ever made and can complete step (4) in 30 seconds, but no miner is ever going to waste 30 seconds of mining power. So to cover the 30 seconds while step (4) is run, the miner has step (2) in place.
Also keep in mind that the pool miners call for the new block asynchronously. So they will hammer the mining server for (2) when a new block shows up, but once they get (2) they will be much more casual about polling the mining server for (4). This generates a latency between when the mining server knows (4) and when the pool miners know (4).
So yeah... the first few minutes after the last block is formed, I'd expect empty block templates to be the norm
- Set your update polling interval for your pool miners to check for new templates periodically
- Create an empty block template linking to the latest received block
- Post (2) for consumption by any miner that calls for a new template
- Generate a second (full) block template from mempool linking to the latest received block
- Post (4) for consumption by any miner that calls for a new template
- If no new block comes in, jump to (3)
- If a new valid block comes in, jump to (2)
The reason why step (2) exists is that step (4) can take MINUTES.... Now we all hope that the machine running getblocktemplate is the biggest fastest meanest manchine ever made and can complete step (4) in 30 seconds, but no miner is ever going to waste 30 seconds of mining power. So to cover the 30 seconds while step (4) is run, the miner has step (2) in place.
Also keep in mind that the pool miners call for the new block asynchronously. So they will hammer the mining server for (2) when a new block shows up, but once they get (2) they will be much more casual about polling the mining server for (4). This generates a latency between when the mining server knows (4) and when the pool miners know (4).
So yeah... the first few minutes after the last block is formed, I'd expect empty block templates to be the norm
Empty blocks are common, and the reasons have been explained many times... And Chinese miners won't change their behavior just by reading these threads, they'll only change if they keep going on the wrong fork 
OROBTC, if you consider Chinese pools "unclean players", then I think Eligius also fits in your "unclean players" list...
OROBTC, if you consider Chinese pools "unclean players", then I think Eligius also fits in your "unclean players" list...
...
Of the last six blocks won, here is some info (11:29 PM US ET) on four of them:
364629 1 tx F2Pool
364632 1 tx F2Pool
364633 1 tx BW.COM
364634 1 tx AntPool
Granted, these were in pretty quick succession, but still.
I don't mine. But if I did, I would go with a clean player like Eligius. Hardware owners may wish to join "the good guys".
Of the last six blocks won, here is some info (11:29 PM US ET) on four of them:
364629 1 tx F2Pool
364632 1 tx F2Pool
364633 1 tx BW.COM
364634 1 tx AntPool
Granted, these were in pretty quick succession, but still.
I don't mine. But if I did, I would go with a clean player like Eligius. Hardware owners may wish to join "the good guys".
There is nothing wrong with empty blocks.
The miners can chose to include transaction or not.
The block reward is given to the miner who can solve the block, it has nothing to do with the acutal transaction. The motivation to include transactions would be to get a slightly higher reward as a result of the tx fees.
My understanding is that by NOT confirming they have a very small time advantage to try and win the block
That's the reason. A block that small will spread way faster through all the nodes and will get faster >50% of the nodes to accept this block as the valid new one. A 1MB block will have a hard time getting through the nodes because he is so big. So he has a chance of getting orphaned, which means he dies while a small, fast blocks survives and becomes the official new block.
...
I stand corrected then, knightdk, thank you.
I read on another thread a day or two ago that Eligius and *someone else* played better with others, hence did not abuse the system.
I stand corrected then, knightdk, thank you.
I read on another thread a day or two ago that Eligius and *someone else* played better with others, hence did not abuse the system.
I don't mine. But if I did, I would go with a clean player like Eligius. Hardware owners may wish to join "the good guys".
Eligius still produces its share of empty blocks. It has a smaller hashrate, so there are less of them.
I don't mine. But if I did, I would go with a clean player like Eligius. Hardware owners may wish to join "the good guys".
Eligius still produces its share of empty blocks. It has a smaller hashrate, so there are less of them....
el kaka22 (great name,
)
My understanding is that by NOT confirming they have a very small time advantage to try and win the block (miner's fees in toto are much lower than block award).
But, I hear you. Such greed by the Chinese miners (when most others apparently do confirm) ought to be seen as very irresponsible. With luck, hardware owners will direct their hashes elsewhere.
It saves time because they don't need to go through the mempool and get transactions to add. Also because they always start out with empty blocks. Preventing this would require a code rewrite which takes time and by the time that the code is done and tested, the attack is over and it is pointless. el kaka22 (great name,
)My understanding is that by NOT confirming they have a very small time advantage to try and win the block (miner's fees in toto are much lower than block award).
But, I hear you. Such greed by the Chinese miners (when most others apparently do confirm) ought to be seen as very irresponsible. With luck, hardware owners will direct their hashes elsewhere.
...
el kaka22 (great name,)
My understanding is that by NOT confirming they have a very small time advantage to try and win the block (miner's fees in toto are much lower than block award).
But, I hear you. Such greed by the Chinese miners (when most others apparently do confirm) ought to be seen as very irresponsible. With luck, hardware owners will direct their hashes elsewhere.
EDIT: There are over 26,250 transactions waiting to be confirmed, 71 MB....
el kaka22 (great name,
)My understanding is that by NOT confirming they have a very small time advantage to try and win the block (miner's fees in toto are much lower than block award).
But, I hear you. Such greed by the Chinese miners (when most others apparently do confirm) ought to be seen as very irresponsible. With luck, hardware owners will direct their hashes elsewhere.
EDIT: There are over 26,250 transactions waiting to be confirmed, 71 MB....
...
Of the last six blocks won, here is some info (11:29 PM US ET) on four of them:
364629 1 tx F2Pool
364632 1 tx F2Pool
364633 1 tx BW.COM
364634 1 tx AntPool
Granted, these were in pretty quick succession, but still.
I don't mine. But if I did, I would go with a clean player like Eligius. Hardware owners may wish to join "the good guys".
So why they don't confirm some unconfirmed txs and get some fees Of the last six blocks won, here is some info (11:29 PM US ET) on four of them:
364629 1 tx F2Pool
364632 1 tx F2Pool
364633 1 tx BW.COM
364634 1 tx AntPool
Granted, these were in pretty quick succession, but still.
I don't mine. But if I did, I would go with a clean player like Eligius. Hardware owners may wish to join "the good guys".
Even if they confirm a 1 satoshi fee tx will get them 1 satoshi more. Plus there are 67k txs in the mempool. I don't see a point that why not take some of the transactions
...
Of the last six blocks won, here is some info (11:29 PM US ET) on four of them:
364629 1 tx F2Pool
364632 1 tx F2Pool
364633 1 tx BW.COM
364634 1 tx AntPool
Granted, these were in pretty quick succession, but still.
I don't mine. But if I did, I would go with a clean player like Eligius. Hardware owners may wish to join "the good guys".
Of the last six blocks won, here is some info (11:29 PM US ET) on four of them:
364629 1 tx F2Pool
364632 1 tx F2Pool
364633 1 tx BW.COM
364634 1 tx AntPool
Granted, these were in pretty quick succession, but still.
I don't mine. But if I did, I would go with a clean player like Eligius. Hardware owners may wish to join "the good guys".
Jump to: