ASICBOOST Aftermath: What Now Must Be Done?

[ebliever]

For those who've missed it:

https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2017-April/013996.html

https://bitcoinmagazine.com/articles/mining-manufacturer-blocking-segwit-benefit-asicboost/

Summary of the situation: Bitmain has been producing and mining with ASICs using an exploit that gives them a dramatic advantage over competing miners. This exploit is incompatible with Segwit, so Bitcoin Unlimited was created to gin up the appearance of public opposition to Core and place monopoly control over bitcoin mining and the protocol in the hands of Bitmain and associates.

The exploit also incentivizes miners to mine empty blocks, which has aggravated the TX bottleneck situation and contributed to high fees.

While they have been exposed, it is not clear to me that there is any reason they will be shut down in the near term. In other words a conspiracy has been unmasked, but not defeated.

ISSUE: ASICBOOST provides a ~30% mining advantage over non-ASICBOOST miners.

NEAR-TERM RISKS:
1. BitMain keeps mining empty blocks. TX fees remain elevated.
2. BU keeps blocking Segwit from reaching 95% consensus. TX capacity remains throttled.
3. By extension, the Lightning Network continues to be blocked/delayed/impaired insofar as SW provides the best implementation for it.

LONG-TERM RISKS:
1. Bitmain runs all opposition out of business and assumes monopoly control over mining.

COUNTERMEASURES:
1. Patent infringement? This is outside of the bitcoin communities' control, but one countermeasure is if Bitmain is infringing on the original ASICBOOST patent and it can be upheld in a Chinese court. I'm not optimistic here but it has to be listed as a possibility.
2. UASF for Segwit/Gregory Maxwell's countermeasures (see link above). A UASF introducing Segwit would force Bitmain to either mine BTC while abandoning their ASICBOOST exploit, or hardfork away with their own altcoin.
3. ? Other options?

LONG-TERM COUNTERMEASURE:
This episode has taken the issue of mining centralization from a relaxed discussion about future risks to bitcoin to an imminent crisis. BU was close to 50% hashrate in the weeks leading up to this expose. A craftier pair than Wu and Ver might pull off a coup next time. Long term no cryptocurrency can survive and thrive in an environment where monopoly control is threatened or possible.

At the same time I want to emphasize that a raw majority of miners did not join BU and should not be unfairly penalized for their actions. They've made tremendous investments in specialized hardware and should not be bankrupted in a rash reaction to this situation.

Fortunately, the ROI on bitcoin mining hardware is pretty short - less than a year from all I've seen. The means the Core development community can, after due process, provide a roadmap covering several years. One that leads back to decentralization of mining. Thus I urge Core to:

1. Provide a roadmap away from the current ASIC-based paradigm back to mining with desktop-based hardware, over a timeframe that allows the mining community to maintain profitability (without the pressure of Bitmain's advantage weighing on them, through SW deployment).

2. Formalize a public policy that such a decentralized mining paradigm will be maintained.  Either by altering algorithms or through other mitigating actions to defeat any ASIC or other specialized hardware implementation that significantly centralizes mining.

Just my 2 satoshis,
ebliever

[Carlton Banks]

1. Provide a roadmap away from the current ASIC-based paradigm back to mining with desktop-based hardware, over a timeframe that allows the mining community to maintain profitability (without the pressure of Bitmain's advantage weighing on them, through SW deployment).

Fully in agreement. We may have averted a real disaster for Bitcoin in the shape of BU, but the plans for the next coup have already been made public, and all the blocksize obsessives on the forums have started the cycle all over again.

The honest miners should not be penalised for maintaining their decorum faced with adversity, but the risk of mining centralisation is still too significant. A staged move away from SHA-2 hashing must happen sooner rather than later to mitigate the risk. Miners can take the opportunity to halt their development of mining ASICs, and concentrate on making the transition to CPU mining, there's no reason why they cannot still compete in the marketplace using their other marginal advantages, which will still exist after chaging the PoW algorithm.

2. Formalize a public policy that such a decentralized mining paradigm will be maintained.  Either by altering algorithms or through other mitigating actions to defeat any ASIC or other specialized hardware implementation that significantly centralizes mining.

Isn't this the same as point 1, though? If the Bitcoin developers write the code, that is the policy, and it couldn't be any more clear written as code. In cryptographic networks like Bitcoin, the code is the policy, the code is the law.


Further to this, it's possible that UASF might not even be needed. Bitmain's loss in hashrate share resulting from gmaxwell's neutering of ASIC Boost might be enough to force Bitmain's hand. Segwit signalling might rise to above 50% from removing ASIC Boost's viability, and 50% BIP acceptance has a tendency to move to 95% very quickly, it makes miners and pools nervous that they may start losing block solution races when 51% threshold is passed.

[franky1]

or blockstream can just admit that segwit isnt perfect by admitting that its 'fixes' are not guaranteed and that it does nothing to stop native spammers etc.

and for blockstream to reset their own snobbery and try to fix their own internal issues and try something the community can and will happily accept

however using PoW nukes, mandatory activations, fee discount bribes and deadlines, bypassing node consensus.. looks very desperate. they are too blind to ven ask why should they need to resort to such tactics..

[ebliever]

Isn't this the same as point 1, though? If the Bitcoin developers write the code, that is the policy, and it couldn't be any more clear written as code. In cryptographic networks like Bitcoin, the code is the policy, the code is the law.

Thanks for your own comments and insights. My last 2 points just distinguish between (1) making the fix to the current problem and (2) making it clear that similar actions will be taken in the future to prevent the problem from recurring.

(I'm a quality engineer, and this breakdown just reflects our habitual thinking as we write corrective action reports and answer a checklist of questions about countermeasures and preventing recurrence.)

[chek2fire]

or blockstream can just admit that segwit isnt perfect by admitting that its 'fixes' are not guaranteed and that it does nothing to stop native spammers etc.

and for blockstream to reset their own snobbery and try to fix their own internal issues and try something the community can and will happily accept

however using PoW nukes, mandatory activations, fee discount bribes and deadlines, bypassing node consensus.. looks very desperate. they are too blind to ven ask why should they need to resort to such tactics..

within the recent Bitmain expose you only see blockstream? lol
hahahahahahahahahahahahahahahahaha

[X7]

or blockstream can just admit that segwit isnt perfect by admitting that its 'fixes' are not guaranteed and that it does nothing to stop native spammers etc.

and for blockstream to reset their own snobbery and try to fix their own internal issues and try something the community can and will happily accept

however using PoW nukes, mandatory activations, fee discount bribes and deadlines, bypassing node consensus.. looks very desperate. they are too blind to ven ask why should they need to resort to such tactics..

So we are discussing an ASICBOOST exploit by BITMAIN and you point to BLOCKSTREAM?

Not only is this irrelevant, but block stream hires a MINORITY of core devs. What are you even on about dude?

[Scott J]

Long term no cryptocurrency can survive and thrive in an environment where monopoly control is threatened or possible.

IOTA gets around this centralisation problem by moving the PoW burden from miners and onto the users (PoW to make a transaction). Miners can't be a problem if they don't exist.

This would probably be far too big a change for Bitcoin going forward though.

[lurker10]

Summary of the situation: Bitmain has been producing and mining with ASICs using an exploit that gives them a dramatic advantage over competing miners.

Welcome to capitalism.
Why is this a big deal?

[Carlton Banks]

 

My last 2 points just distinguish between (1) making the fix to the current problem and (2) making it clear that similar actions will be taken in the future to prevent the problem from recurring.

(I'm a quality engineer, and this breakdown just reflects our habitual thinking as we write corrective action reports and answer a checklist of questions about countermeasures and preventing recurrence.)

I did understand the difference.

Did you see what I mean when I said "the code is the policy". If the code is written in a way that equates to the policy, what is the purpose of writing public policy statements? The code is the most meaningful public policy statement there could be, written words in human language are essentially promises. Spoken promises can be broken, the programming logic cannot.

Hence, the code is a more valuable statement of intent than any written statement could ever be. I realise I'm talking down your vocation, but perhaps your skills would be better transferred to some other area as we move into a true cypher punk era. I guess we'll have to wait and see, but I suspect cypher punk philosophy has the potential to propagate faster than many might predict.

[aso118]

Can't the exploit be fixed without Segwit getting activated?
This can be taken care of in the next upgrade, without getting into the Segwit vs BU debate.

[cellard]

1. Provide a roadmap away from the current ASIC-based paradigm back to mining with desktop-based hardware, over a timeframe that allows the mining community to maintain profitability (without the pressure of Bitmain's advantage weighing on them, through SW deployment).

Fully in agreement. We may have averted a real disaster for Bitcoin in the shape of BU, but the plans for the next coup have already been made public, and all the blocksize obsessives on the forums have started the cycle all over again.

The honest miners should not be penalised for maintaining their decorum faced with adversity, but the risk of mining centralisation is still too significant. A staged move away from SHA-2 hashing must happen sooner rather than later to mitigate the risk. Miners can take the opportunity to halt their development of mining ASICs, and concentrate on making the transition to CPU mining, there's no reason why they cannot still compete in the marketplace using their other marginal advantages, which will still exist after chaging the PoW algorithm.

2. Formalize a public policy that such a decentralized mining paradigm will be maintained.  Either by altering algorithms or through other mitigating actions to defeat any ASIC or other specialized hardware implementation that significantly centralizes mining.

Isn't this the same as point 1, though? If the Bitcoin developers write the code, that is the policy, and it couldn't be any more clear written as code. In cryptographic networks like Bitcoin, the code is the policy, the code is the law.


Further to this, it's possible that UASF might not even be needed. Bitmain's loss in hashrate share resulting from gmaxwell's neutering of ASIC Boost might be enough to force Bitmain's hand. Segwit signalling might rise to above 50% from removing ASIC Boost's viability, and 50% BIP acceptance has a tendency to move to 95% very quickly, it makes miners and pools nervous that they may start losing block solution races when 51% threshold is passed.

Don't be so sure. Roger Ver is still a big whale and so is Jihad Wu itself. The bribing will continue and the propaganda war never ends, plus the miners that neutrally think segwit is not a good idea because they don't want to lower fees since it will give them less money.

Looks like it's a tie and BTC will remain as it is, unless UASF kicks in and I think it's not that easy.

LTC will get segwit for sure and BTC will still struggle, im buying more LTC right now.

[jonald_fyookball]

For those who've missed it:

https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2017-April/013996.html

https://bitcoinmagazine.com/articles/mining-manufacturer-blocking-segwit-benefit-asicboost/

Summary of the situation: Bitmain has been producing and mining with ASICs using an exploit that gives them a dramatic advantage over competing miners. This exploit is incompatible with Segwit, so Bitcoin Unlimited was created to gin up the appearance of public opposition to Core and place monopoly control over bitcoin mining and the protocol in the hands of Bitmain and associates.

"exploit" ... anyone is free to design their own ASIC... Given the nature of Bitcoin, I would say it would be wrong to enforce the patent and not let miners compete freely but that is for the courts to decide.

BU has nothing to do with this -- we simply need bigger blocks, this has been going back to classic, xt, etc... 1mb is BS, stop listening to core propaganda please.

[chek2fire]

For those who've missed it:

https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2017-April/013996.html

https://bitcoinmagazine.com/articles/mining-manufacturer-blocking-segwit-benefit-asicboost/

Summary of the situation: Bitmain has been producing and mining with ASICs using an exploit that gives them a dramatic advantage over competing miners. This exploit is incompatible with Segwit, so Bitcoin Unlimited was created to gin up the appearance of public opposition to Core and place monopoly control over bitcoin mining and the protocol in the hands of Bitmain and associates.

"exploit" ... anyone is free to design their own ASIC... Given the nature of Bitcoin, I would say it would be wrong to enforce the patent and not let miners compete freely but that is for the courts to decide.

BU has nothing to do with this -- we simply need bigger blocks, this has been going back to classic, xt, etc... 1mb is BS, stop listening to core propaganda please.

is strange that the same ppl that use this shady tactics are the same that not only support BU but and every community toxic split attempt.
Their reasons is now very clear because and for and and for most ppl was a puzzle why they do it.
The reason is that Bitmain plan is to centralised mining.
Everyone that believe that Bitmain want big block is simple a stupid. They only to create a stall situation and nothing more. this is their plan and until now it works great.

[ebliever]

Summary of the situation: Bitmain has been producing and mining with ASICs using an exploit that gives them a dramatic advantage over competing miners.

Welcome to capitalism.
Why is this a big deal?

Did you read the rest of the OP?

Capitalism thrives on a free market. There is a reason monopolies are prohibited by capitalist nations where feasible, and a monopoly is where things are headed at this point.

Granted, we can just say "There are a  thousand Alts out there, we'll just switch to one of them." And that would be a valid, free-market response to a miner hegemony taking control of bitcoin and squeezing bitcoin users with maximized fees. But that would create tremendous turmoil and financial losses by many bitcoin investors and supporters, negative press that would make Mt. Gox look like a jaywalking incident, and damage the whole cryptocurrency movement for years.

[ebliever]

You're still buying into the propaganda.   Hang Jihan from the highest tree if you want... i don't care -- We simply need bigger blocks.

Oh hogwash. We need increased transaction capacity. We have SW and LN, and you BU shills have been fighting progress for all you are worth.

It's like listening to someone keep saying that we need a 2nd horse on the horse cart, when semi trucks with diesel engines are available.

[chek2fire]

https://www.reddit.com/r/Bitcoin/comments/63otrp/gregory_maxwell_major_asic_manufacturer_is/dfvwga4/

The suspicion was motivated by observing that SegWit was very likely incompatible with an optimized implementation if it--- which happened by chance, basically I was saying "to block asicboost the network could do something like <xxx>" and then I realized the words I was saying were basically part of the SegWit design.
With this in mind many otherwise hard to explain facts clicked into place-- e.g. aggressive attacks on Bitcoin Core that started after the proposal of asicboost, arguments against segwit that seemed to make no sense, advocacy for "hardfork segwit"... and this justified further investigation.
I hoped to find a test that would conclusively show which blocks were using it, but this doesn't appear to be possible.
More recently the "extension block + lightning" discussions have also stricken people as inexplicable (since many thought that segwit was being opposed because it potentially facilitated off-chain transactions)-- but they also fit.

[franky1]

lets see what the maths really tells us

just done some quick maths


* stats at time of post

hmm
looks like BTCC and F2pool are the ones making more blocks than their hash %
not the other way round

*for those wishing to question the numbers

i would have expected antpool to have a block % of something in the 40's while having hash in the 30's if all this gmaxwell PoW propaganda was real
definitely not less than their hash%

oh well gmaxwell debunked.

kind of funny how many times gmaxwells announcement didnt name the pool and how many times gmaxwell uses the word "they could" rather then "they are"

A month ago I was explaining the attack on Bitcoin's SHA2 hashcash which
is exploited by ASICBOOST and the various steps which could be used to
block it in the network if it became a problem.

While most discussion of ASICBOOST has focused on the overt method
of implementing it, there also exists a covert method for using it.

An incompatibility would go a long way to explain some of the
more inexplicable behavior from some parties in the mining
ecosystem so I began looking for supporting evidence.

Reverse engineering of a particular mining chip has demonstrated
conclusively that ASICBOOST has been implemented
in hardware.

Due to a design oversight the Bitcoin proof of work function has a potential
attack which can allow an attacking miner to save up-to 30% of their energy
costs (though closer to 20% is more likely due to implementation overheads).

..
just to let the script writers twist it into "its an attack, bomb them bomb them bomb them"

P.S
gotta laugh that when its an exploit.. he words it as exploiting [adambacks] hashcash.
but when its a bug he calls it an bitcoin proof of work oversight....

he is too far deep inside his bosses pocket

[chek2fire]

You're still buying into the propaganda.   Hang Jihan from the highest tree if you want... i don't care -- We simply need bigger blocks.

Oh hogwash. We need increased transaction capacity. We have SW and LN, and you BU shills have been fighting progress for all you are worth.

It's like listening to someone keep saying that we need a 2nd horse on the horse cart, when semi trucks with diesel engines are available.

SG not activated...wont be... has less support from miners than BU.

That's like me saying:  "I need a car"... and you saying "well you can have this 1942 ford pickup in my driveway that barely hits 45 mph... but you
can only have it if 95% of the neighborhood agrees..but right now you have to walk."

LN still needs big blocks to work (it says so in the whitepaper) .....why can't you see that Core is obstructing common sense scaling?  It's so plain and obvious.

I don't care about BU.... I really don't ... calling me a "BU shill" just shows me you've swallowed the propaganda.

BU is just one way to get bigger blocks.  There's classic, coreEC, flexcap, other ABC proposals, extension blocks..etc... segwit is
the least palatable on-chain scaling , its minimal, and miners don't support it.

what propaganda you talk about now? Is not propaganda that a single Miner with the most hashrate attack the majority of bitcoin developers and propose a crap flaw code that twice crashed in two months only by a simple call script?
Is not propaganda that miners refuse to enable an over a year tested solution again from the majority of bitcoin developers?
and is propaganda everything else?
Look man. There are three types of ppl with your opinion. Payed shadow puppets, tech blind, or trolls. You can choose where you fit there.

[chek2fire]

lets see what the maths really tells us

just done some quick maths


* stats at time of post

hmm
looks like BTCC and F2pool are the ones making more blocks than their hash %
not the other way round

*for those wishing to question the numbers

i would have expected antpool to have a block % of somethin in the 40's while having hash in the 30's if all this gmaxwell PoW propaganda was real

this post is very clear to anyone that you not even understand what we talk about here.... what a crap
what has to do this with ASICBOOST? Ver is You?

[lurker10]

Summary of the situation: Bitmain has been producing and mining with ASICs using an exploit that gives them a dramatic advantage over competing miners.

Welcome to capitalism.
Why is this a big deal?

Did you read the rest of the OP?

Capitalism thrives on a free market. There is a reason monopolies are prohibited by capitalist nations where feasible, and a monopoly is where things are headed at this point.

A monopoly is only bad when it is given unfair advantage via regulations and privileges made law by government.
When a monopoly is naturally formed, it is a product of free market.
Who can stop any miner from implementing the same "exploit" in their ASIC? Nobody. Is there a law against that? No. Then it's fair game.