Topic: it is Core, not Bitman blocking segwit - page 6. (Read 5420 times)

To you it may be the current issue.  

To me, its a non-issue.

To me, Blocksize debate is the only issue
that's important.  Everything else is a distraction.

It is not about Maxwell and SegWit.
It is about restricting certain avenues of development because a miner took a risk and lost.
You are advocating that this miner is "too big to fail" and so the community needs to bail them out.

The community should not be subjected to this miner's bad business decisions.
That is anti-Bitcoin and anti-Satoshi. We are not Ethereum.
Optimizations are acceptable, "zero-day" like exploits in PoW is playing fast and loose.

It needs to be patched. In time the mining community will likely agree.

"it is Core, not Bitman blocking segwit"

Do you not see how ridiculous this sounds to even the most minimally informed person? You're way past the point where the best thing you could do is just... stop writing. Just stop. Just walk away from the keyboard and tell Roger/Jihan you can't do this any more.

prevent improvements?

seriously..!!
then gmaxwell should could program new code that does work. rather than be adement that the code should not change one bit but hardware should..

blaming hardware is no excuse for sloppy code.
especially when the code doesnt live up to ANY of its promises anyway.

i can think of many ways to fix it in software but after dealing with his snobbery last year where it took a good 6 months for him to accept the issue was an issue and then without admission change a few things round.. screw it if im gonna help him again

I agree. ASICBoost was not intended to prevent SegWit and I don't think
anyone is reasonably thinking that. The issue is that allowing ASICBoost to
continue and be used through the mining community makes development in
certain aspects very restrictive and harder to do. Satohi's Fraud Proofs
relied on the Coinbase Ref idea and expected that Coinbase refs will be used
in the future to help improve the network. Widespread ASICBoost use will
prevent that avenue we have, potentially indefinitely.



I disagree. The use of the ASICBoost mechanism prevents improvements to
the Bitcoin network. It is not an optimization of the mining process. In fact, it
uses a short cut or loophole. When GPUs became ASICs that is somewhat acceptable,
since that was a true optimization in technology. ASICBoost exploits a failing in PoW
which creates non-mining and far reaching problems for the community.

It is one thing to create new mining devices to beat your competitors, its another to
find "zero-days" which could be used to beat your competitor as well as prevent
further expansion of the bitcoin code in ways that even Satoshi envisioned.
That is the difference here. One is natural, the other is a bug exploit.

Aren't miners now all mining with ASICs? then how can they vote for something that will defuse them makes them useless?
Changing POW of bitcoin to what exactly? what can we mine bitcoin with after such a change? let them cheat in mining we
just need to get involved in trading bitcoin, it's not like common folks can jump in mining with their laptops or desktops
after POW changes anyways.

what i mean is asicboost is not a "segwit attack/exploit"
asicboost is just an efficiency gain.. end off..

its greg denying his bad code, by using his snobbery to hide that his code is just not compatible with hardware of 2 years ago


so they said no to segwit because they knew segwit had a incompatibility bug with efficient hardware..
EG like saying you cant use bitcoin with an ATI GPU in 2011, because of an issue that was added in code in 2011 that only works with geforce GPU's

do you blame ATI or the code writers for not making mining easy.

No. The mining community agreed not to use ASICBoost for many different reasons.
In fact, today, no miner that I am aware of openly admits using it, since its taboo.

In addition, using that on a large scale would hinder many importation developments
that can not be implemented as cleanly as with new Coinbase references.

The Miners can not dictate the direction of efficient new features and optimization.
They need to perform the full PoW and that is all. If they make custom hardware
that cheats the PoW, that was a risky move that the rest of the community does
not need to be subjected to, neither the innocent users or innocent miners.

That was their financial gamble. Some dice rolls are one too many.
In time, we will learn what the true scale and impact this issue is.

Bitmain in their recent statement admitted that there chips have ASICBoost
built in, but have not used them on the main chain. Are you calling Bitmain
timetravelers then? Obviously not. This means they had them for a longer
time than people assumed.

When SegWit was released, Bitmain and it's pool, Antpool, did not yet know
that the Coinbase reference is added to with anchor codes. When they learned
it did, they likely began to oppose SegWit on those grounds secretly but
declared publicly it was because there would be no 2MB blocks.

This is a reasonable premise and is being investigated.

If Maxwell is a liar or spreading FUD as you and Bitman believe, that will be
determined in time. All things come out in the wash in time.

asic boost has been around longer.
easy fix.. make software that doesnt hurt the hardware.. not the other way round.


cant blame the hardware of 2 years ago for software problems of 6 months go

thats like saying ATI exploited bitcoin by being better at GPU mining in 2011-2013 compared to Geforces poor efforts

that like blaming raspberry Pi1 for not validating signatures fast, and so libsecp256k1 had to be invented because raspberry Pi was attacking bitcoin

Those aren't irrelevant. This may be the real issue at hand.

The significance is allowing miners to continue using such an
exploit prevents Bitcoin from upgrading in different ways.
Many of Satoshi's ideas of extra features are dependent on
Coinbase references. ASICBoost prevents any new Coinbase
references.

Blocksize debate is irrelevant in this current issue. Using
ASICBoost prevents many forms of development that helps
scaling and non-scaling issues. It is potentially more serious
than anything else currently.

If Miners are allowed to continue with ASICBoost, Bitcoin
development ends in many fundamental ways.

You have sided with one miners profits over the whole future
of the Bitcoin community. You do not know what you are
hitching yourself to.

so your saying they did go back in time and made S9chips to include asic boost to attack segwit..
wow amazing.

or logic dictates that segwit wasnt programmed well to actually work with asic efficiency boosts.
Hint: which is easier
1. go back in time by 2 years and undo hardware changes to meet temporary software requirements just 6 months ago, just for some blockstreamer
2. get the centralist blockstream team to now change their code seeing as its only been in public for 6 months and not even active code yet

calling an efficiency boost an attack, even though the efficiency boost existed prior to the code that is supposedly being attacked.. has been the funniest twist of words by greg this month.

anyway


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%

  
Let's not get distracted here with irrelevant details.

Even IF Bitmain was using ASICboost, (no evidence they are)
and even IF that was illegal (its not since they hold the Chinese patent),
so what?
  
Point is, they support segwit (if done with big block) , and also extension
blocks.  (neither is compatible with asicboost btw). 
Core won't support anything but their roadmap.   So, which party is
unilaterally preventing a scaling solution...honestly?

Ach, I really don't give a shit any more. Everyone from every angle is becoming equally tedious.

I'm off to invent a tiny brain that lives in a box that out mines everything else on Earth by a factor of 1 billion and then I'll toss a coin as to which version I prefer.

Some clarification on those "facts":

1. Jihan supported before understanding the final SegWit Implementation code.
After learning the Coinbase references are used, he would obviously retract that support.

2. Extension blocks do not change the Coinbase references, only add new anchor txs.
Those anchor txs should not effect the way ASICBoost works, the way new Coinbase refs do.

3. This comment is jumping the gun. In depth investigations begin now.

4. This data set should be larger and go farther back in time. Likely prior to the patent dates.
The ASICBoost could be throttled from time to time to prevent obvious indicators.

5. Fees are irrelevant here. The exploit centers around gaining more block rewards.
In a future with less block reward and more fees, this exploit is worthless.

6. We need independent verification, which will begin now.


It would likely be best for you to stop quoting Alex.BTC since it is obvious that he
is not interested in learning anything, but perpetuating the obfuscations.
In time, all will be revealed. This is a new development and nicely explains many
previously unknown factors. If motive needed to be determined, this would be a
reasonable assumption to investigate further, and so the community will.

Your adherents to your viewpoints in light of new information still being gathered, is telling.

no great reveal...just more propaganda.

Erm, I thought the great reveal of the last couple of days was Segwit nullifying Bitmain's advantage. I don't know how recently this advantage was discovered and put into action but that explains plenty of recent attitudes.

I'm not saying that SegWit is not compatible with ASICs.
I'm saying that ASICs with ASICBoost is not compatible with new Coinbase references.
And that may be the true reason why SegWit blocking Miners are actually against SegWit.
If SegWit could be rewritten without the Coinbase change, they may be fine with SegWit.

So, miners that have ASICBoost can not accept new features that add data to the Coinbase.
This is the theory. Nothing to do with SegWit or ASICs specifically. It is about an addition
that causes the full PoW work to be performed into the future and renders ASICBoost like
designs and utilization worthless.

your assuming that bitmain read october 2016's segwit code
and then time travels back to 2015 and made blueprints for a chip that could attack segwit..

..
more likely sceneario.. segwit is not ASIC compatible .. not the other way round.(unless time travel is possible)

end result
0.13.x versions have issues with ASIC and no longer going to be segwit compatible ..
time to wait a month and let 0.14.1 to be released and let blockstream restart their nodecount that will work with ASICS

My point isn't so much about the miners, but Core.

Core won't even compromise to 2mb/segwit, which
probably could get consensus.  
 

Miners don't want 'just segwit'.  They want a HF and blocksize increase. 
Easy for everyone to point the finger at everyone else and say the
other one is the obstructionist, but before Segwit came along,
there was discussion of 4mb vs 8mb (then later 2mb) and that
was the debate. 

Core created the entire 'HF are bad' narrative and then segwit
out of nowhere.

At least miners are willing to compromise, and that is the difference.
Core is not.