Wondering out loud: Which should Chinese miners support - Core, Classic or another?

[sturle]

I hope they are going to upgrade to 0.12 before they release.  0.12 has lots of improvements, e.g. much faster IBD, data usage limiting, the long awaited RBF feature, etc.  It will be hard to convince people to downgrade.

The first version will be 0.11.2 with a 2MB block size. It may very well be the final version if Core decides to wise up.

Doesn't seem to be a serious attempt then.  Are the Toomins too incompetent to merge the improvements from 0.12?

The choice of 0.11.2 was a conscious decision in order to make it perfectly clear what this release of Classic is about.

They wanted everything else than the 2MB block size limit increase to be uncontroversial. This is best achieved with 0.11.2. Not everyone are as enthusiastic about RBF as you are.

RBF is optional in 0.12.  Users don't have to use RBF if they don't want to.

And all the other improvements?  Is e.g. data limiting to satisfy network data quotas controversial?  Or the speed improvements?

Of course it does. I agree that there needs to be more focus on efficiency, but my view (and that of quite a few others as well) is that we can't let the cow starve while the grass is growing. And this seems to be the main point of contention.

Letting the cow die in a civil war isn't much better.

... the civil war which was started in order to save the cow from starving?

The cow isn't starving.  The blocks aren't full, except during spam attacks which would have filled blocks of any size.  Not even then, since most miners don't produce full blocks under any circumstance.  Normally, e.g. now, it is far between full blocks.  Some miners are quite good at filtering the spam, and confirm even 0-fee transactions in the first block they make even during the worst spam storms.  Doubling the size of the what field won't help a bit, when you don't do sh*t about the real problem.  Swarm of grasshoppers and birds eating everyting they get.

Are we going to have an equally fruitful debate about the word "full" as we had about "released"?

Scroll around and look at the recent blocks:

https://tradeblock.com/bitcoin/

https://blockchain.info/

There's not a lot of room for growth.

By the way, nice metaphor for adoption: "Swarm of grasshoppers and birds eating everyting they get."

Adoption?  It that what you call transaction spam?  Do you think 2 MB blocks will reduce the spam in any way?  The blocks will be just as full.

I just had a look.  Of the last 8 blocks, only one is close to full.  1.8 kB left in that one, which should allow for about 9 more non-segwit transactions.  The rest have plenty room.

That is just silly-talk. Last I checked all the blocks were 900kB+. Ideally it should be no more than 10-20% full in order for the network to be ready to absorb more users.

Tell that to the spammers.  There are plenty enough transactions coming to fill 20 MB blocks today, if you want all the junk to litter your disk and network, and get most full nodes off the network.  2 MB blocks will be just as full as 1 MB blocks.  The only limiting factor right now is the fact that most miners demand a minimum amount of fee per KB to avoid building too large blocks, or block some of the spam, because larger blocks take longer to spread.  My mempool has hundreds of megabytes of spam in it.

The blocks which are completely full usually contain a lot of dust transactions, i.e. outputs which cost more in fees tp spend than the value of the output, which won't get mined using a default configurations.  Some miners happily prefer those transactions as long as they pay more in fees.  This is abuse, imho.

Do you have any data to back up your claim of the prevalence of dust spam in blocks now?

I can only give examples.  Here is one from the latest block from F2Pool, one of the pools including dust as long as it pays enough fees.  0.0000082 BTC sent, paying a fee of 0.0001 BTC.

And what do you think of this?  Does it look like economic activity or spam to you?  If you follow the transactions, you see they are just spending fees to send very small amounts (0.0001 BTC) in circles.

This won't decrease, no matter how much block space you throw at it.  Larger blocks will only make room for more transactions of this kind.  You could have gigabytes of block space, and there will always be another spammer spending a few bitcents on fees to "prove" it is not enough, claiming that "bitcoin is dead" and arguing for centralization, control and censorship.

Here is a nice infographic which explains why segwit is a much better idea than some panic fork blocksize war.

In case you don't remember:

I hope Core replies with a 2MB/Segwit hard fork. It seems like a clever piece of technology. But I wasn't talking about Segwit. Segwit will only get us so far.

I answered.  Segwit will get us further.  It makes room for more transactions than 2MB blocks will, is easier and safer to deploy, contains lots of other improvements, and scales much better in the future.

[1714878896]

1714878896

[1714878896]

1714878896

[1714878896]

1714878896

[Fatman3001]

Are we going to have an equally fruitful debate about the word "full" as we had about "released"?

Scroll around and look at the recent blocks:

https://tradeblock.com/bitcoin/

https://blockchain.info/

There's not a lot of room for growth.

By the way, nice metaphor for adoption: "Swarm of grasshoppers and birds eating everyting they get."

Adoption?  It that what you call transaction spam?  Do you think 2 MB blocks will reduce the spam in any way?  The blocks will be just as full.

I just had a look.  Of the last 8 blocks, only one is close to full.  1.8 kB left in that one, which should allow for about 9 more non-segwit transactions.  The rest have plenty room.

That is just silly-talk. Last I checked all the blocks were 900kB+. Ideally it should be no more than 10-20% full in order for the network to be ready to absorb more users.

Tell that to the spammers.  There are plenty enough transactions coming to fill 20 MB blocks today, if you want all the junk to litter your disk and network, and get most full nodes off the network.  2 MB blocks will be just as full as 1 MB blocks.  The only limiting factor right now is the fact that most miners demand a minimum amount of fee per KB to avoid building too large blocks, or block some of the spam, because larger blocks take longer to spread.  My mempool has hundreds of megabytes of spam in it.

The blocks which are completely full usually contain a lot of dust transactions, i.e. outputs which cost more in fees tp spend than the value of the output, which won't get mined using a default configurations.  Some miners happily prefer those transactions as long as they pay more in fees.  This is abuse, imho.

Do you have any data to back up your claim of the prevalence of dust spam in blocks now?

I can only give examples.  Here is one from the latest block from F2Pool, one of the pools including dust as long as it pays enough fees.  0.0000082 BTC sent, paying a fee of 0.0001 BTC.

And what do you think of this?  Does it look like economic activity or spam to you?  If you follow the transactions, you see they are just spending fees to send very small amounts (0.0001 BTC) in circles.

This won't decrease, no matter how much block space you throw at it.  Larger blocks will only make room for more transactions of this kind.  You could have gigabytes of block space, and there will always be another spammer spending a few bitcents on fees to "prove" it is not enough, claiming that "bitcoin is dead" and arguing for centralization, control and censorship.

So the answer is: NO

Of course there is dust spam, there's just nobody who's been able to show any convincing data on the matter. If someone could show me that the average block size without genuinely useless (that would have to be proven too) dust spam was along the lines of 150-200kB I might give it more weight.

[hdbuck]

There is no data regarding the relevance of inflating the blocksize either.

Besides the "an internet page on average is 2MB.. let's do 20MB.. because Moore ad infinitum" -USGavin

[sAt0sHiFanClub]

Bitcoin has been designed to work fine without incoming connections.  A non-upgraded node can make outgoing connections to anyone.  All nodes.  Upgraded or not.  Upgraded nodes will accept incoming connections from anyone.  Upgraded or not.  The only change he is suggesting, is to make sure upgraded nodes only make outgoing connections to other upgrading nodes.  My own node currently has 8 outgoing and 104 incoming connections.  Even if my node is upgraded, and all my outgoing connections are to segwit nodes, there is nothing whatsoever stopping non-upgraded nodes to connect to my node, and nothing is splitting the network in any way imaginable.

How does a peer to peer network work if all nodes only make outgoing connections? How does that work exactly?

That wouldn't work, of course.  Another reason not to worry.  Unfortunately many nodes are behind NAT and/or restrictive firewalls, and cannot accept incoming connections.  Nodes which are on a public IP, and accept incoming connections, often have a hundred or more.

Logical fallacy. Dont make a statement like "Bitcoin has been designed to work fine without incoming connections" and then pretend to educate about port forwarding.

I can see why you dont understand the idea of how segwit nodes selectively picking connections will split the network.

[sturle]

And what do you think of this?  Does it look like economic activity or spam to you?  If you follow the transactions, you see they are just spending fees to send very small amounts (0.0001 BTC) in circles.

This won't decrease, no matter how much block space you throw at it.  Larger blocks will only make room for more transactions of this kind.  You could have gigabytes of block space, and there will always be another spammer spending a few bitcents on fees to "prove" it is not enough, claiming that "bitcoin is dead" and arguing for centralization, control and censorship.

So the answer is: NO

Please answer my question above.  I think it shows very clearly how large the spam problem is by one example, showing just a small part of this spammers transactions on the blockchain.  How would larger blocks solve this problem?

Of course there is dust spam, there's just nobody who's been able to show any convincing data on the matter. If someone could show me that the average block size without genuinely useless (that would have to be proven too) dust spam was along the lines of 150-200kB I might give it more weight.

You know very well that task is impossible to prove.  Segwit increases the capacity and make secure 0-conf off-chain transactions very easy to do.  Spamming the blockchain to make normal transactions unreliable, increase centralization, promote censorship and split the community would no longer work.

[sturle]

Bitcoin has been designed to work fine without incoming connections.  A non-upgraded node can make outgoing connections to anyone.  All nodes.  Upgraded or not.  Upgraded nodes will accept incoming connections from anyone.  Upgraded or not.  The only change he is suggesting, is to make sure upgraded nodes only make outgoing connections to other upgrading nodes.  My own node currently has 8 outgoing and 104 incoming connections.  Even if my node is upgraded, and all my outgoing connections are to segwit nodes, there is nothing whatsoever stopping non-upgraded nodes to connect to my node, and nothing is splitting the network in any way imaginable.

How does a peer to peer network work if all nodes only make outgoing connections? How does that work exactly?

That wouldn't work, of course.  Another reason not to worry.  Unfortunately many nodes are behind NAT and/or restrictive firewalls, and cannot accept incoming connections.  Nodes which are on a public IP, and accept incoming connections, often have a hundred or more.

Logical fallacy. Dont make a statement like "Bitcoin has been designed to work fine without incoming connections" and then pretend to educate about port forwarding.

OK, I'll find a baby spoon:
Bitcoin has been designed in such a way that running a node without incoming connections work fine.

Of course you need at least one node on the network which is able to accept incoming transactions.  Preferably more than one.

I can see why you dont understand the idea of how segwit nodes selectively picking connections will split the network.

I have tried ELI5.  I just don't understand why you can't understand it.  Please try to read again.

As I stated above, the bitcoin network depends on at least one node able to accept incoming connections.  That node will accept incoming connections from all other nodes.  Upgraded or not.  It doesn't matter.  The only way to split this, is if the single node is not upgraded.  In that case all upgraded nodes will be isolated until at least one of them is able to accept incoming connections.  But this isn't likely to happen, is it?  Another way to split this network in the case above, would be to switch off the only node on the entire network which accept incoming connections.  In that case nobody can connect to other nodes, and every node is isolated.

Fortunately in the real world there is more than one node on the network able to accept incoming connections, so this will not be a problem.  If at least one of the nodes which accept incoming transactions is upgraded, there won't be a split.  Both upgraded and non-upgraded nodes will have a common connection point there.  When 95% of the hashrate is on upgraded nodes already (otherwise they wouldn't be able to create upgraded blocks), we can safely assume the upgraded nodes are well connected.

[Fatman3001]

And what do you think of this?  Does it look like economic activity or spam to you?  If you follow the transactions, you see they are just spending fees to send very small amounts (0.0001 BTC) in circles.

This won't decrease, no matter how much block space you throw at it.  Larger blocks will only make room for more transactions of this kind.  You could have gigabytes of block space, and there will always be another spammer spending a few bitcents on fees to "prove" it is not enough, claiming that "bitcoin is dead" and arguing for centralization, control and censorship.

So the answer is: NO

Please answer my question above.  I think it shows very clearly how large the spam problem is by one example, showing just a small part of this spammers transactions on the blockchain.  How would larger blocks solve this problem?

All the transactions I looked at were 373 bytes. Assuming they're all 373 bytes they would amount to 4.4MB since early Nov 2015. This information tells me nothing.

Of course there is dust spam, there's just nobody who's been able to show any convincing data on the matter. If someone could show me that the average block size without genuinely useless (that would have to be proven too) dust spam was along the lines of 150-200kB I might give it more weight.

You know very well that task is impossible to prove.

And that is a real problem. But proper estimates would help.

Segwit increases the capacity and make secure 0-conf off-chain transactions very easy to do.  Spamming the blockchain to make normal transactions unreliable, increase centralization, promote censorship and split the community would no longer work.

Cool! Throw it in with a 2MB hard fork. Personally I guess I would prefer it with BIP101, but that doesn't seem very politically realistic.

[sturle]

And what do you think of this?  Does it look like economic activity or spam to you?  If you follow the transactions, you see they are just spending fees to send very small amounts (0.0001 BTC) in circles.

This won't decrease, no matter how much block space you throw at it.  Larger blocks will only make room for more transactions of this kind.  You could have gigabytes of block space, and there will always be another spammer spending a few bitcents on fees to "prove" it is not enough, claiming that "bitcoin is dead" and arguing for centralization, control and censorship.

So the answer is: NO

Please answer my question above.  I think it shows very clearly how large the spam problem is by one example, showing just a small part of this spammers transactions on the blockchain.  How would larger blocks solve this problem?

All the transactions I looked at were 373 bytes. Assuming they're all 373 bytes they would amount to 4.4MB since early Nov 2015. This information tells me nothing.

If that tells you nothing, I feel genuinely sorry for you.  You are blind by choice, or someone gave you glasses with coloured glass.  Some people are investing a lot of money in splitting the community, and you won't accept the evidence.  This is a small drop in the ocean of spam.  And it proves my point that spam fills the blocks.  Most of the transactions are unconfirmed, and as soon as they are, the coins are sent on and spent on more spam.  Actually the coins are usually just sent without waiting for a confirmation, creating large trees of unconfirmed transactions.

Of course there is dust spam, there's just nobody who's been able to show any convincing data on the matter. If someone could show me that the average block size without genuinely useless (that would have to be proven too) dust spam was along the lines of 150-200kB I might give it more weight.

You know very well that task is impossible to prove.

And that is a real problem. But proper estimates would help.

More qualified people than me have estimated the amount of spam to at least 3/4 of all transactions, but the number is of course not certain.  There are "mixing services" as well generating a ton of transactions.  CT and payment channels will help get rid of those, I hope.  Segwit makes it all much easier.

Segwit increases the capacity and make secure 0-conf off-chain transactions very easy to do.  Spamming the blockchain to make normal transactions unreliable, increase centralization, promote censorship and split the community would no longer work.

Cool! Throw it in with a 2MB hard fork. Personally I guess I would prefer it with BIP101, but that doesn't seem very politically realistic.

A hard fork won't solve any real world problems.  Just create more, by splitting the community in a brutal way.  Many large and rich businesses, e.g. R3,  want to destroy Bitcoin and create a new blockchain where they are in control, and they may have found the cheapest way to do it.  Now they are in a hurry, since they know for a fact that segwit and the possibilities enabled by segwit will make a 51% attack their only option, and their investments until now will be lost.  Their best hope is control over a hard fork by incompetent people, and a community hostile to progress.

[Carlton Banks]

A hard fork won't solve any real world problems.  Just create more, by splitting the community in a brutal way.  Many large and rich businesses, e.g. R3,  want to destroy Bitcoin and create a new blockchain where they are in control, and they may have found the cheapest way to do it.  Now they are in a hurry, since they know for a fact that segwit and the possibilities enabled by segwit will make a 51% attack their only option, and their investments until now will be lost.  Their best hope is control over a hard fork by incompetent people, and a community hostile to progress.

+21 million

fatman, you really ought to reconsider throwing your weight behind blocksize hard forks, it makes you look like a massive shill for R3CEV

[hdbuck]

A hard fork won't solve any real world problems.  Just create more, by splitting the community in a brutal way.  Many large and rich businesses, e.g. R3,  want to destroy Bitcoin and create a new blockchain where they are in control, and they may have found the cheapest way to do it.  Now they are in a hurry, since they know for a fact that segwit and the possibilities enabled by segwit will make a 51% attack their only option, and their investments until now will be lost.  Their best hope is control over a hard fork by incompetent people, and a community hostile to progress.

+21 million

fatman, you really ought to reconsider throwing your weight behind blocksize hard forks, it makes you look like a massive shill for R3CEV

weight? what weight??!

[Carlton Banks]

fatman, you really ought to reconsider throwing your weight behind blocksize hard forks, it makes you look like a massive shill for R3CEV

weight? what weight??!

I'm assuming that "3001" refers to what the bathroom scales tell him  

[Fatman3001]

And what do you think of this?  Does it look like economic activity or spam to you?  If you follow the transactions, you see they are just spending fees to send very small amounts (0.0001 BTC) in circles.

This won't decrease, no matter how much block space you throw at it.  Larger blocks will only make room for more transactions of this kind.  You could have gigabytes of block space, and there will always be another spammer spending a few bitcents on fees to "prove" it is not enough, claiming that "bitcoin is dead" and arguing for centralization, control and censorship.

So the answer is: NO

Please answer my question above.  I think it shows very clearly how large the spam problem is by one example, showing just a small part of this spammers transactions on the blockchain.  How would larger blocks solve this problem?

All the transactions I looked at were 373 bytes. Assuming they're all 373 bytes they would amount to 4.4MB since early Nov 2015. This information tells me nothing.

If that tells you nothing, I feel genuinely sorry for you.  You are blind by choice, or someone gave you glasses with coloured glass.  Some people are investing a lot of money in splitting the community, and you won't accept the evidence.  This is a small drop in the ocean of spam.  And it proves my point that spam fills the blocks.  Most of the transactions are unconfirmed, and as soon as they are, the coins are sent on and spent on more spam.  Actually the coins are usually just sent without waiting for a confirmation, creating large trees of unconfirmed transactions.

Keep your eyes on the ball, my reply was in relation to this question :

The blocks which are completely full usually contain a lot of dust transactions, i.e. outputs which cost more in fees tp spend than the value of the output, which won't get mined using a default configurations.  Some miners happily prefer those transactions as long as they pay more in fees.  This is abuse, imho.

Do you have any data to back up your claim of the prevalence of dust spam in blocks now?

I can only give examples.

Your example doesn't prove that the blocks are full of spam. It just likely proves that there is spam on the network.

We already agree about that:

Of course there is dust spam, there's just nobody who's been able to show any convincing data on the matter. If someone could show me that the average block size without genuinely useless (that would have to be proven too) dust spam was along the lines of 150-200kB I might give it more weight.

You know very well that task is impossible to prove.

And that is a real problem. But proper estimates would help.

More qualified people than me have estimated the amount of spam to at least 3/4 of all transactions, but the number is of course not certain.  There are "mixing services" as well generating a ton of transactions.  CT and payment channels will help get rid of those, I hope.  Segwit makes it all much easier.

I've read that as well somewhere. And I read a lot of people who disagreed with him. To be fair, what they all had in common was that they were just pulling numbers out of their exhaust ports.

Segwit increases the capacity and make secure 0-conf off-chain transactions very easy to do.  Spamming the blockchain to make normal transactions unreliable, increase centralization, promote censorship and split the community would no longer work.

Cool! Throw it in with a 2MB hard fork. Personally I guess I would prefer it with BIP101, but that doesn't seem very politically realistic.

A hard fork won't solve any real world problems.  Just create more, by splitting the community in a brutal way.  Many large and rich businesses, e.g. R3,  want to destroy Bitcoin and create a new blockchain where they are in control, and they may have found the cheapest way to do it.  Now they are in a hurry, since they know for a fact that segwit and the possibilities enabled by segwit will make a 51% attack their only option, and their investments until now will be lost.  Their best hope is control over a hard fork by incompetent people, and a community hostile to progress.

Ok, now I get it.

[Poloniex Matthew]

0.00099 is far above the dust threshold.  It is 513 satoshis or thereabouts.  The monetary value, even assuming 0 txfee, is much less than one cent.

[sAt0sHiFanClub]

Bitcoin has been designed to work fine without incoming connections.  A non-upgraded node can make outgoing connections to anyone.  All nodes.  Upgraded or not.  Upgraded nodes will accept incoming connections from anyone.  Upgraded or not.  The only change he is suggesting, is to make sure upgraded nodes only make outgoing connections to other upgrading nodes.  My own node currently has 8 outgoing and 104 incoming connections.  Even if my node is upgraded, and all my outgoing connections are to segwit nodes, there is nothing whatsoever stopping non-upgraded nodes to connect to my node, and nothing is splitting the network in any way imaginable.

How does a peer to peer network work if all nodes only make outgoing connections? How does that work exactly?

That wouldn't work, of course.  Another reason not to worry.  Unfortunately many nodes are behind NAT and/or restrictive firewalls, and cannot accept incoming connections.  Nodes which are on a public IP, and accept incoming connections, often have a hundred or more.

Logical fallacy. Dont make a statement like "Bitcoin has been designed to work fine without incoming connections" and then pretend to educate about port forwarding.

OK, I'll find a baby spoon:
Bitcoin has been designed in such a way that running a node without incoming connections work fine.

Of course you need at least one node on the network which is able to accept incoming transactions.  Preferably more than one.

I can see why you dont understand the idea of how segwit nodes selectively picking connections will split the network.

I have tried ELI5.  I just don't understand why you can't understand it.  Please try to read again.

As I stated above, the bitcoin network depends on at least one node able to accept incoming connections.  That node will accept incoming connections from all other nodes.  Upgraded or not.  It doesn't matter.  The only way to split this, is if the single node is not upgraded.  In that case all upgraded nodes will be isolated until at least one of them is able to accept incoming connections.  But this isn't likely to happen, is it?  Another way to split this network in the case above, would be to switch off the only node on the entire network which accept incoming connections.  In that case nobody can connect to other nodes, and every node is isolated.

Fortunately in the real world there is more than one node on the network able to accept incoming connections, so this will not be a problem.  If at least one of the nodes which accept incoming transactions is upgraded, there won't be a split.  Both upgraded and non-upgraded nodes will have a common connection point there.  When 95% of the hashrate is on upgraded nodes already (otherwise they wouldn't be able to create upgraded blocks), we can safely assume the upgraded nodes are well connected.

You can ELI5 all you like. But you are only explaining it to yourself at this stage.

Bitcoin has been designed to work fine without incoming connections

[sturle]

You can ELI5 all you like. But you are only explaining it to yourself at this stage.

I know.  You clearly shut your head off, and I think everyone else on Bitcointalk understood my explanation long ago.

[xyzzy099]

0.00099 is far above the dust threshold.  It is 513 satoshis or thereabouts.  The monetary value, even assuming 0 txfee, is much less than one cent.

The current dust limit is 2730 satoshis.  It used to be 546 before all the 'stress testing'.

[sturle]

0.00099 is far above the dust threshold.  It is 513 satoshis or thereabouts.  The monetary value, even assuming 0 txfee, is much less than one cent.

The current dust limit is 2730 satoshis.  It used to be 546 before all the 'stress testing'.

Correct.  It depends on relayfee.  Now that mempool limiting has been implemented (in 0.12) it is safer to lower relayfee again.

[classicsucks]

A hard fork won't solve any real world problems.  Just create more, by splitting the community in a brutal way.  Many large and rich businesses, e.g. R3,  want to destroy Bitcoin and create a new blockchain where they are in control, and they may have found the cheapest way to do it.  Now they are in a hurry, since they know for a fact that segwit and the possibilities enabled by segwit will make a 51% attack their only option, and their investments until now will be lost.  Their best hope is control over a hard fork by incompetent people, and a community hostile to progress.

QED. That is a solid theory, but to quote Harvey Keitel, "Let's not start suckin' each other dicks quite yet".

The purpose of the spamming could also be to keep nodes and their network links at peak capacity. It appears that there is a Bitcoin NETWORK LIMIT to the total transactions processed (NOT JUST the blocksize), and there is a system for FILTERING transactions.  I believe that these limits are being approached commonly. Logically, people who have their transactions processed over others' are in a preferred position and therefore can CONTROL THE PRICE to some degree...

So maybe these hundreds of 5k-15k BTC transactions per day are big players transferring coin to one another with their resources pooled - perhaps they're just buying coin back and forth between hundreds of accounts with no net profit or loss?  Such transactions, especially if prioritized higher, could have a dramatic influence on price. Logically, people invested in mining hardware would have an incentive to use such a system to keep the BTC price up... Would such actors want a blocksize increase? Or better control over spam transaction filtering? Or even a covert transaction prioritization system or network?

ALL JUST A THEORY... come at me, trolls.