Author

Topic: Gold collapsing. Bitcoin UP. - page 134. (Read 2032266 times)

legendary
Activity: 1764
Merit: 1002
July 08, 2015, 03:08:33 PM
the ongoing spam shows clearly that all full nodes are handling the traffic just fine. We were told that they would crash and burn from overloaded memory if we were foolhardy enough to implement Gavin's arbitrary, rushed, poorly-researched plan for 20MB blocks.

^Fixted for you.

The real shame is that they've been forced to do ask the validation and store it in mempool waiting for blocks that never come because of the 1MB cap. What a waste but it shows that the capacity of the network is far higher than we've been told.

If you look at the stream  of full blocks going by, to me I see miners begging to be able to process bigger blocks to clear their mempools. That would force the spammers losses and eventually kill him. 

Mempool is full of tx waiting to be (slowly, thanks to completely unoptimized Createnewblock) processed into new blocks.

The blocks are then propagated (slowly thanks to limited/expensive upstream bandwidth), and (eventually, thanks to slow ECDSA and quadratic scaling) validated by the receivers.

Validation is not done in mempool, as gmax and others have tried to explain to you, without success and at the cost of your remaining credibility.


The network will upgrade as a result of new user growth that the spammer can no longer disrupt. Meanwhile, the spammer gets taken out to the wood shed and gets raped. She won't be coming back.

The network does not distinguish between new users and old users.  More users will include more spammers.

Wow, you've really gone off the deep end today.  Consider a vacation, or therapy.  It's not healthy for a man of your considerable age/education/income/stature to talk like a poorly disciplined young teen on their X-Box Halo chat.



All that slowness you're referring to results in the spam being deleted after 24 hours allowing recycling of the fees while still disrupting usage.
legendary
Activity: 2156
Merit: 1072
Crypto is the separation of Power and State.
July 08, 2015, 02:49:57 PM
the ongoing spam shows clearly that all full nodes are handling the traffic just fine. We were told that they would crash and burn from overloaded memory if we were foolhardy enough to implement Gavin's arbitrary, rushed, poorly-researched plan for 20MB blocks.

^Fixted for you.

The real shame is that they've been forced to do ask the validation and store it in mempool waiting for blocks that never come because of the 1MB cap. What a waste but it shows that the capacity of the network is far higher than we've been told.

If you look at the stream  of full blocks going by, to me I see miners begging to be able to process bigger blocks to clear their mempools. That would force the spammers losses and eventually kill him. 

Mempool is full of tx waiting to be (slowly, thanks to completely unoptimized Createnewblock) processed into new blocks.

The blocks are then propagated (slowly thanks to limited/expensive upstream bandwidth), and (eventually, thanks to slow ECDSA and quadratic scaling) validated by the receivers.

Validation is not done in mempool, as gmax and others have tried to explain to you, without success and at the cost of your remaining credibility.


The network will upgrade as a result of new user growth that the spammer can no longer disrupt. Meanwhile, the spammer gets taken out to the wood shed and gets raped. She won't be coming back.

The network does not distinguish between new users and old users.  More users will include more spammers.

Wow, you've really gone off the deep end today.  Consider a vacation, or therapy.  It's not healthy for a man of your considerable age/education/income/stature to talk like a poorly disciplined young teen on their X-Box Halo chat.

legendary
Activity: 1414
Merit: 1000
July 08, 2015, 02:47:39 PM
Spammer only spends money 1x during lifetime, but miner has to keep bigger disk-space forever (2 hard disc consume more electricity than single one)

OK.  Provide me with your estimates for the following (and explain how you arrived at your numbers) and I'll update my table using your numbers:

1.  The cost per node to store 1 GB of additional blockchain data for 5 years, assume the outputs are spent.
2.  The cost per node to store 1 GB of additional blockchain data for 5 years, assuming the outputs are unspent.



I'm sorry, I have no estimates. I know that bloat-chain is not solution  There are better solutions how to organize blockchain.

My full node just crashed ... all data are lost ... downloading 6 years 25 weeks.  ... hmm I hope that it is only me and there will remain some copy of blockchain. :-)  (not joking)
legendary
Activity: 1764
Merit: 1002
July 08, 2015, 02:37:49 PM
Whole network must upgrade. So spammer spend $1200 but bitcoin network 10x more on upgrade,

You must have quoted my post before I updated it with the new table to estimate the cost to the network that you were referring.  I think "10x more" is an overstatement.  It still looks like it costs the spammer more than the network as a whole [blue column] if we write the spam to the blockchain, even though I think the relevant cost is the cost per node:



Spammer only spends money 1x during lifetime, but miner has to keep bigger disk-space forever (2 hard disc consume more electricity than single one)

So what, without a cap miner will get paid to do it. And profit and get stronger which is the last thing the spammer wants especially since he can no longer disrupt the user growth which will, in the meantime, lead to squaring of the networks value (price) according to Metcalfe Law.
legendary
Activity: 1162
Merit: 1007
July 08, 2015, 02:34:14 PM
Spammer only spends money 1x during lifetime, but miner has to keep bigger disk-space forever (2 hard disc consume more electricity than single one)

OK.  Provide me with your estimates for the following (and explain how you arrived at your numbers) and I'll update my table using your numbers:

1.  The cost per node to store 1 GB of additional blockchain data for 5 years, assume the outputs are spent.
2.  The cost per node to store 1 GB of additional blockchain data for 5 years, assuming the outputs are unspent.
hero member
Activity: 644
Merit: 504
Bitcoin replaces central, not commercial, banks
July 08, 2015, 02:34:08 PM
What network disruption are we talking about here?

legendary
Activity: 1764
Merit: 1002
July 08, 2015, 02:30:38 PM
Why would the cost to a particular node depend on the number of nodes?

lol, do we have single centralized server ?

We don't.
But why should a single node owner pay for the storage used by every other node in the network, instead of bearing just its share ?
Because that's what it looks you're implying...




Whole network must upgrade. So spammer spend $1200 but bitcoin network 10x more on upgrade,

The network will upgrade as a result of new user growth that the spammer can no longer disrupt. Meanwhile, the spammer gets taken out to the wood shed and gets raped. She won't be coming back.
legendary
Activity: 1414
Merit: 1000
July 08, 2015, 02:29:36 PM
Whole network must upgrade. So spammer spend $1200 but bitcoin network 10x more on upgrade,

You must have quoted my post before I updated it with the new table to estimate the cost to the network that you were referring.  I think "10x more" is an overstatement.  It still looks like it costs the spammer more than the network as a whole [blue column] if we write the spam to the blockchain, even though I think the relevant cost is the cost per node:



Spammer only spends money 1x during lifetime, but miner has to keep bigger disk-space forever (2 hard disc consume more electricity than single one)
legendary
Activity: 1764
Merit: 1002
July 08, 2015, 02:26:12 PM
A block cap doesn't discourage spam; it encourages it. That's because not only is it a target to cause congestion but the attacker knows he can congest nodes that process the spam immediately upon receipt and then store it in memory waiting for a block that never comes. Since unconfirmed TX's get deleted from mempool after 24 hours, the spam costs him nothing while at the same time disrupting user experience and growth.

What a deal. 
nby
newbie
Activity: 27
Merit: 0
July 08, 2015, 02:25:11 PM
Why would the cost to a particular node depend on the number of nodes?

lol, do we have single centralized server ?

We don't.
But why should a single node owner pay for the storage used by every other node in the network, instead of bearing just its share ?
Because that's what it looks you're implying...




Whole network must upgrade. So spammer spend $1200 but bitcoin network 10x more on upgrade,

On the other hand:
Spammer spends 0 (no fees) while whole network spends K for bandwidth

Edit: I agree in full with Peter R and what his table reports.
legendary
Activity: 1162
Merit: 1007
July 08, 2015, 02:22:22 PM
Whole network must upgrade. So spammer spend $1200 but bitcoin network 10x more on upgrade,

You must have quoted my post before I updated it with the new table to estimate the cost to the network that you were referring.  I think "10x more" is an overstatement.  It still looks like it costs the spammer more than the network as a whole [blue column] if we write the spam to the blockchain, even though I think the relevant cost is the cost per node:

legendary
Activity: 1414
Merit: 1000
July 08, 2015, 02:17:08 PM
Why would the cost to a particular node depend on the number of nodes?

lol, do we have single centralized server ?

We don't.
But why should a single node owner pay for the storage used by every other node in the network, instead of bearing just its share ?
Because that's what it looks you're implying...




Whole network must upgrade. So spammer spend $1200 but bitcoin network 10x more on upgrade,
legendary
Activity: 1414
Merit: 1000
July 08, 2015, 02:15:19 PM


Why would the cost to a particular node depend on the number of nodes?
lol, do we have single centralized server ?

OK, it sounds like you're looking at the costs to the network as a whole, rather than the cost to an individual node.  Indeed this cost will be higher (cost/node * number of nodes).  

However, the table I posted specifically addressed the cost per node.  So...which is the important cost?  In my opinion, it's the cost per node.  As a node operator or as a miner, I only care how much it costs me.  If my actions can effectively charge the spammer $1200 while costing me a few pennies, I'll do it!  Whether a thousand or a million other nodes also contribute a few pennies to penalize the spammer makes no difference to me.  

How are you looking at the issue?

Why do you think that there is only 1 spammer ? In reality there can be much more (100x ?) spammers than miners.  
nby
newbie
Activity: 27
Merit: 0
July 08, 2015, 02:12:57 PM
Why would the cost to a particular node depend on the number of nodes?

lol, do we have single centralized server ?

We don't.
But why should a single node owner pay for the storage used by every other node in the network, instead of bearing just its share ?
Because that's what it looks you're implying...

legendary
Activity: 1764
Merit: 1002
July 08, 2015, 02:12:30 PM
Spam tx's are by definition economic transactions someone finds inappropriate, and individual judgments calls are not part of the protocol, that's bitcoin

Miners can choose what to include in the blocks they build.

anyway to say 6 years after bitcoin has grown that storing transactions and blocks for ever should change is not keeping with the idea that is Bitcoin.  

OK. I'm not sure anyone suggested otherwise.

Don't forget that with pruning you'll be able to trim your node if you want while turning the full validation over to specialized nodes.

I think bandwidth is a bigger issue than storage costs. Pruning does nothing for bandwidth.

Yes, but the ongoing spam shows clearly that all full nodes are handling the traffic just fine. We were told that they would crash and burn from overloaded memory. Not true.

The real shame is that they've been forced to do ask the validation and store it in mempool waiting for blocks that never come because of the 1MB cap. What a waste but it shows that the capacity of the network is far higher than we've been told.

If you look at the stream  of full blocks going by, to me I see miners begging to be able to process bigger blocks to clear their mempools. That would force the spammers losses and eventually kill him. 
legendary
Activity: 2156
Merit: 1072
Crypto is the separation of Power and State.
July 08, 2015, 02:09:38 PM
I think bandwidth is a bigger issue than storage costs. Pruning does nothing for bandwidth.

Thanks to the end of the economic crisis, benevolent cash-flush corporations will SoonTM begin rolling out fiber to every home so every pig farmer in Moldova/Greece/Florida/Ukraine/etc. can binge watch Marco Polo in 1080p on Netflix.  (According to Gavin; YMMV.)

Just as soon as their stocks start trading again....WAIT FOR IT...any second now...  Embarrassed    Huh   Embarrassed
legendary
Activity: 1162
Merit: 1007
July 08, 2015, 02:06:17 PM
Why would the cost to a particular node depend on the number of nodes?
lol, do we have single centralized server ?

OK, it sounds like you're looking at the costs to the network as a whole, rather than the cost to an individual node.  Indeed this cost will be higher (cost/node * number of nodes).  

However, the table I originally posted specifically addressed the cost per node. Here's a new one where I've estimated the cost to the network (blue column):



So...which is the important cost?  In my opinion, it's the cost per node.  As a node operator or as a miner, I only care how much it costs me.  If my actions can effectively charge the spammer $1200 while costing me a few pennies, I'll do it!  Whether a thousand or a million other nodes also contribute a few pennies to penalize the spammer makes no difference to me.  

How are you looking at the issue?
legendary
Activity: 2044
Merit: 1005
July 08, 2015, 01:56:01 PM
Spam tx's are by definition economic transactions someone finds inappropriate, and individual judgments calls are not part of the protocol, that's bitcoin

Miners can choose what to include in the blocks they build.

anyway to say 6 years after bitcoin has grown that storing transactions and blocks for ever should change is not keeping with the idea that is Bitcoin.  

OK. I'm not sure anyone suggested otherwise.

Don't forget that with pruning you'll be able to trim your node if you want while turning the full validation over to specialized nodes.

I think bandwidth is a bigger issue than storage costs. Pruning does nothing for bandwidth.
legendary
Activity: 1414
Merit: 1000
July 08, 2015, 01:55:58 PM
Why would the cost to a particular node depend on the number of nodes?

lol, do we have single centralized server ?
legendary
Activity: 2156
Merit: 1072
Crypto is the separation of Power and State.
July 08, 2015, 01:49:55 PM
Don't forget that with pruning you'll be able to trim your node if you want while turning the full validation over to specialized nodes.

Yes, let's all prune trustlessness, the raison d'etre of bitcoin, from our nodes for the sake of adoption at any cost.

/African kid fetish
Jump to: