New transaction malleability attack wave? Another stresstest?

[monsterer]

This stress-test wasn't direct attempt to prove anything.
I do not how to explain it. It is like a chess-game.
You can donate a chess piece to your opponent or make a nonclear turn to win a game.

What makes you think that trustless systems always have to tend towards centralisation?

[amaclin]

What makes you think that trustless systems always have to tend towards centralisation?

Because it is more economically reasonable solution in long term.
The centralized system takes less energy. Always. Point. No exceptions. Look around. Look to yourself.
Decentralized system either takes more energy or less secure in long term.

[mikeymillie]

Besides this, Satoshi seems to have had quite a few reasons to develop
Bitcoin such as fleeing from banks, creating a trustless system, etc.

OK. So my reason is to protect your life savings from this ponzi scheme called bitcoin
I want to prove that decentralized trustless system can not exists in long term.
It either transforms to centralized system or loses its security.

>>> It either transforms to centralized system or loses its security.

This is correct.  I disagree about the "Ponzi" comment, but that is irrelevant really.  

I wonder if anyone in Bitcoin has ever thought about how this pure decentralized system has no counterpart in nature?  None.  

A game:   cite of an example where the universe has evolved a natural system exhibiting coordination of behavior by a full decentralization of group consensus, and I will tell you why it is wrong.  

Bitcoin isn't supernatural, it's not special just because humans made it and can "outsmart" the constraints of the medium they exist in.  

Energy loss, signal propagation limits, information entropy,  nature itself abhors decentralization (of the Bitcoin sort) and will tend to centralize due to self-signalling feedback effects, or lose signal integrity due to entropy.  Consensus, when confronted with the hard limits of signalling in the medium it is sustained, must specialize by losing least significant inputs or decompose into fragments.

No getting around this.  Bitcoin cannot "win" the battle to scale while remaining decentralized; there is no possible win condition unless the criteria for "decentralization" are loosened (which, abstractly, is what larger blocks and overlay/sidechains both do).

[amaclin]

Bitcoin cannot "win" the battle to scale while remaining decentralized;

But centralized bitcoin - is nonsense. We already have FED.

I disagree about the "Ponzi" comment, but that is irrelevant really.

I mean that all earnings in this system can be get only from the money from next members.

[monsterer]

A game:   cite of an example where the universe has evolved a natural system exhibiting coordination of behavior by a full decentralization of group consensus, and I will tell you why it is wrong.  

The starfish.

https://en.wikipedia.org/wiki/Starfish

While a starfish lacks a centralized brain, it has a complex nervous system with a nerve ring around the mouth and a radial nerve running along the ambulacral region of each arm parallel to the radial canal. The peripheral nerve system consists of two nerve nets: a sensory system in the epidermis and a motor system in the lining of the coelomic cavity. Neurons passing through the dermis connect the two.[28] The ring nerves and radial nerves have sensory and motor components and coordinate the starfish's balance and directional systems.[5] The sensory component receives input from the sensory organs while the motor nerves control the tube feet and musculature. The starfish does not have the capacity to plan its actions. If one arm detects an attractive odour, it becomes dominant and temporarily over-rides the other arms to initiate movement towards the prey. The mechanism for this is not fully understood.

[mikeymillie]

Bitcoin cannot "win" the battle to scale while remaining decentralized;

But centralized bitcoin - is nonsense. We already have FED.

I disagree about the "Ponzi" comment, but that is irrelevant really.

I mean that all earnings in this system can be get only from the money from next members.

> But centralized bitcoin - is nonsense. We already have FED.

Not disagreeing here, but centralization and decentralization are not the only states possible.  They are two abstract ideal ends on a continuum.  A fully centralized system doesn't exist; centralization of information asymptotically approaches total singularity, and decentralization likewise approaches total entropy.

> I mean that all earnings in this system can be get only from the money from next members.

I believe what you are trying to make is a practical, moral observation about who benefits from Bitcoin's cash flows. But really this is necessarily true of any system that requires its participants to distinguish between the value of resources outside itself vs inside itself, and who must choose, based on this distinction, whether to move some of their own resources from outside to inside.

[mikeymillie]

A game:   cite of an example where the universe has evolved a natural system exhibiting coordination of behavior by a full decentralization of group consensus, and I will tell you why it is wrong.  

The starfish.

https://en.wikipedia.org/wiki/Starfish

While a starfish lacks a centralized brain, it has a complex nervous system with a nerve ring around the mouth and a radial nerve running along the ambulacral region of each arm parallel to the radial canal. The peripheral nerve system consists of two nerve nets: a sensory system in the epidermis and a motor system in the lining of the coelomic cavity. Neurons passing through the dermis connect the two.[28] The ring nerves and radial nerves have sensory and motor components and coordinate the starfish's balance and directional systems.[5] The sensory component receives input from the sensory organs while the motor nerves control the tube feet and musculature. The starfish does not have the capacity to plan its actions. If one arm detects an attractive odour, it becomes dominant and temporarily over-rides the other arms to initiate movement towards the prey. The mechanism for this is not fully understood.

Good choice.  There are a couple like this, and some better ones, but they are still wrong.

For the starfish has no central nervous system, but it does have a radially centralized body plan - and more subtly, a solid body is an information signalling medium of its own. If one arm pulls, the rest of the creature comes with it.  A brain need not agree to this, or even a network of nerves. The entire physical unity of its mass does the job of "maintaining consensus" about a direction for the creature to move. To lose this consensus (such as arms pulling in opposite directions, or being pulled by outside forces) would be to physically fragment the creature. The nervous system described above is an overlay network which helps prevent this from happening.  

Also, there are natural constraints to how big starfish can become before they cannot sustain physical integrity.  

[amaclin]

Not disagreeing here, but centralization and decentralization are not the only states possible.  They are two abstract ideal ends on a continuum.  A fully centralized system doesn't exist; centralization of information asymptotically approaches total singularity, and decentralization likewise approaches total entropy.

Who is the winner between two players in a game? When all other factors are the same?
The person, who spends less energy (takes less resources outside the system)

I believe what you are trying to make is a practical, moral observation about who benefits from Bitcoin's cash flows. But really this is necessarily true of any system that requires its participants to distinguish between the value of resources outside itself vs inside itself, and who must choose, based on this distinction, whether to move some of their own resources from outside to inside.

This is very good point.
Let us calculate together the "product" and the "resources" based on principles "outside" and "inside".
What is the product of any transaction system like fiat money and bitcoin?
What banks do produce? What bitcoin system produces?

Banks do not produce money and bitcoin network do not produce bitcoins.
Both systems produce secure transactions.

In this game the winner will be the system, where the cost of resources per one transaction is less (if we talk about the same security)
These systems can co-exist together if one of them is better in security for users and another is better by cost of usage.
Right now bitcoiners think that the cost of usage is small. They are confused the "cost of usage" and "transaction fees".
Transaction fees are small, but the cost of usage is very high.
It can be calculated as ( AmountOfElectricuty + CostOfHardWare ) / NumberOfTransactions

[mikeymillie]

> Who is the winner between two players in a game? When all other factors are the same?

How long is the game?  

> In this game the winner will be the system, where the cost of resources per one transaction is less (if we talk about the same security)

The big catch is that bitcoiners do not recognize security that is based on maintaining a loop of sterile, mechanically rigid, clearly understood consensus rules that they can see in the open, as being something fundamentally comparable to security based on trusting other people to be, in the aggregate, slightly more honest than corrupt when working out of sight.

For bitcoiners, the fact that "decentralized" security costs more is worth it, because in their view it is a better quality product worth paying a premium for.   A decentralization premium.  

[amaclin]

> Who is the winner between two players in a game? When all other factors are the same?
How long is the game?

While both of them able to fight.

For bitcoiners, the fact that "decentralized" security costs more is worth it, because in their view it is a better quality product worth paying a premium for.   A decentralization premium.

The cost is high, but nobody sees it, because the payment is delayed for future. Like in every pyramid scheme.

[monsterer]

The cost is high, but nobody sees it, because the payment is delayed for future. Like in every pyramid scheme.

The cost is subsidised by the block reward.

You are forgetting the cost of 'lost trust' in centralised systems; governments and banks can make terrible decisions which have a huge cost to those relying on those systems. This lost trust may compensate for the increased decentralisation cost in p2p currencies.

[amaclin]

This is the question:

How long is the game?

And this is the answer:

The cost is subsidised by the block reward.

So, the game will be over when the cost is more than block reward.
Do not count the "cost" in dollars. Count it in joules or kwh.

After it we will see decreasing hashrate and ongoing decreasing the security of confirmed transactions.
All the factors which allowed bitcoin system to grow in past will push the system down.

Today you can easy reorganize the blockchain in some dead altcoin with an obsolete asic.
Tomorrow you will be able to reorganize bitcoin blockchain with your currently running asic.
Somebody definitely will "test" is because "he will be able to do it with small efforts"

I repeat: I think not about today. Like a chess-master I think several turns in future.

[btcdrak]

If you are changing bitcoin core in response to this you are likely doing something wrong.

You can simply test anything here on your own too, just use two wallets in regtest mode and sign a transaction twice to get two versions.

Running this attack makes it hard to collect data on which signer software needs to be updated to produce lowS signatures-- which is important for fixing the behavior--, so it would certainly be preferable if it weren't going on. (... not like this thread actually gives a darn about fixing the behavior. )

May I suggest writing to the bitcoin-dev@ mailing list asking wallet authors to check their code behavior? It's not perfect, but, in theory, it's likely to get a chance of being read by the relevant people.

[TooDumbForBitcoin]

This is the question:

How long is the game?

And this is the answer:

The cost is subsidised by the block reward.

So, the game will be over when the cost is more than block reward.
Do not count the "cost" in dollars. Count it in joules or kwh.

After it we will see decreasing hashrate and ongoing decreasing the security of confirmed transactions.
All the factors which allowed bitcoin system to grow in past will push the system down.

Today you can easy reorganize the blockchain in some dead altcoin with an obsolete asic.
Tomorrow you will be able to reorganize bitcoin blockchain with your currently running asic.
Somebody definitely will "test" is because "he will be able to do it with small efforts"

I repeat: I think not about today. Like a chess-master I think several turns in future.

Your December, 2014 "BTC price to King's pawn and $10 value by March 2015" gambit cost you several thousand Grand Master points.  Have you been reading more chess books since then?

[amaclin]

Your December, 2014 "BTC price to King's pawn and $10 value by March 2015" gambit cost you several thousand Grand Master points.  Have you been reading more chess books since then?

OK, bitcoin network is little bit stronger than i thought a year ago
But the physics remain the same.

[monsterer]

So, the game will be over when the cost is more than block reward.

By that time, it is supposed that transaction fees will compensate. No one can predict the future, if they do no problem, if they don't there is a problem.

[btcdrak]

So, the game will be over when the cost is more than block reward.

By that time, it is supposed that transaction fees will compensate. No one can predict the future, if they do no problem, if they don't there is a problem.

Not if blocksize keeps increasing it wont: bitcoin would either have to find an alternative subsidy or die.

[dexX7]

How do I determine whether it is signed with highS or not ?

Step 1: find the signature in the scriptSig

In your example the format is:

Code:

<signature> <pubkey>

To visually explore it, the ASM code or a block explorer of your choice may help. I really like this one:

http://srv1.yogh.io/#tx:id:36d047abcb966f58aa668f050d60254730a3c07c9fd51e869e8b1a773c05d516



Step 2: split the signature into components



Via bitcoin.stackexchange.com

Note that the image also includes the "signature hash type", which is not part of the DER encoding, but usually shown on explorers.

Step 3: check, whether the S value is below the curve order

Compare:

Code:

7fffffffffffffffffffffffffffffff5d576e7357a4501ddfe92f46681b20a1 <- order n of G
6a5611667d1eb147d6a7352cbf37a2ddec8d9b37fcad17a0c9a9c6caff287f24 <- the S value

In this case the S value is indeed smaller than n, and therefore it's "low S".

[amaclin]

By that time, it is supposed that transaction fees will compensate. No one can predict the future, if they do no problem, if they don't there is a problem.

By that time year numbers are all less than 2016.
No one can predict the future.
If last six year numbers were all less that 2016 - there is no problem to think that this is true forever
Do you see logic here? 

[monsterer]

By that time, it is supposed that transaction fees will compensate. No one can predict the future, if they do no problem, if they don't there is a problem.

By that time year numbers are all less than 2016.
No one can predict the future.
If last six year numbers were all less that 2016 - there is no problem to think that this is true forever
Do you see logic here? 

If adoption of bitcoin remains constant (as of now) there needs to be a transaction backlog in order for fees to compensate.