August 1st consideration and important unaswered questions

[cahulahupo]

Edit:  Please look at my response to the first responder for more clarity on the questions posted.  Do not assume I am making some error in logic.  For instance, a reorg is possible on the BIP148 chain with another BIP148 chain which was covertly mined.  This could happen even in September destroying BTC.  Also there is no way to know if ASICBOOST abusers would be profitable without ASICBOOST.  It may be a reasonable assumption but it is an assumption with huge consequences if wrong.


I have some questions regarding August 1st. I have diligently read as much as I can to the point that I would not learn anything new by continuing to read more so I thought it was time to create an account to ask the questions I don't believe have been answered.  I am primarily asking for the purposes of trading safely but these questions go well beyond a speculative nature and so I thought this was the proper place for them.  I appreciate any responses but I want to keep the level of conversation productive and so generalizations are not appreciated.  I have numbered the questions so responders can respond to individual questions and not all of them.  If you aren't an authority or have first hand knowledge of the subject matter please do not respond.


1.  How is it possible that any pro-segwit movement can reach any significant amount of mining power support because in the mining world of perfect competition only those taking advantage of covert ASICBOOST should be able to survive?  Shouldn't all honest miners be losing money in the long run, at least if difficulty adjusted faster to price increases?  Perhaps by August 1st the ASICBOOST miners will have been limited by development time frames but surely if the UASF is not successful by Aug 1st, bitcoin will never be able to adopt segwit with any realistic amount of mining power.  That is because ACISBOOST abusers will not only be self-sufficient but much more flush with cash to reinvest in more hashing power until honest miners are extinct from uncompetitiveness.  Am I correct in my assumptions?

2.  Even if UASF is supported by the community, if LN isn't ready I'm afraid from a practical point of view users will turn to a BU fork because of lower transaction costs even if they themselves think BU is a shit long term scaling solution. If you needed to send a $10 payment and had some money on an exchange, wouldn't you buy the BU coins with your BTC and initiate a $0 fee transfer instead of paying $5 in fees?

3.  If LN isn't instantly deployable how long will it take to be deployed?  Will sites like exchanges which create an enormous amount of on chain transfers be able to instantly reduce their transfer load through LN and if so by how much?

4.  A continuation of question 3, in terms of development time and effort what cost will merchants pay to switch to using LN?  I'm not saying what it will take them to recognize non-segwit transactions on the segwit chain, I know that is trivial.  I am asking how long will it take and how tedious will it be for them and their users to start benefiting from lower transaction costs.

5.  If we agree that the ASICBOOST miners have acted Machiavellian up to this point, why exactly would we ever expect them to honor any agreement where they would sacrifice their profitability?  If they lose the ability to exploit ASICBOOST, that means they could go from being high profit miners to being losers in the mining space no matter what price bitcoin reached. If they no longer have a competitive advantage against other miners, they will lose money regardless of the BTC price.  With this in mind how is a malicious attack on BIP148 different from their secretive ASICBOOST exploitation and the faux politics of certain miners?  Why would they possibly not attack the BIP148 chain if what they have done so far is actually worse?  A long chain reorg that causes user losses would be worse but an above board attack where BIP148 users are not given a false sense of security would not be worse, it would just be a continuation of disruption.

6.  When will ETH reach the bottleneck state that bitcoin is in right now?  Is the potential of an imminent flippening a silly concept due to the similar limitations on ETH?  I would assume that if bitcoin continues to stagnate a statis as seen in punctuated equilibrium states of evolutionary biology is likely to develop as fees increase on ETH.  Any telling when ETH will reach such a bottleneck?  I imagine the first to solve the scaling issue would then dominate unless maybe another solved it shortly after.  If neither BTC nor ETH solves their scaling issue within the next 9 months and LTC has a working LN, I can imagine a LTC market cap of greater than 40billion.

7.  If UASF does not succeed it seems to reason that ETH will certainly scale first due to its more agile development unless its future scaling problems will take much longer than expected to roll out.  Can anyone speak to this?

8.  LTC already has working Segwit but doesn't afaik have LN set up on it.  Why is this?  When will LN be deployable on LTC?

9.  How could a post-UASF chain ever truly know within a reasonable time period that there will not be a reorganization?  If we assume that there isn't any decrease in the ASICBOOST hashing power and that a large amount of malicious miners have not been dormant in anticipation and or are not just coming online can we be confident in the BIP148 chain?  Are those reasonable assumptions to make?

10.  To end off on a note of more personal politics, I think both Core and BU have a terrible actor or two.  Ver is obviously not honest and Theymos who I am aware doesn't work for Blockstream could have done a much better job at constructively moderating /r/bitcoin.  Ad hominem attacks aside, increasing the blocksize is obviously not a long term scaling situation and the schism can only be resolved with having this conflict come to a decisive end so this is the best thing for bitcoin and crypto in general.  It may seem like the fundamentals do not matter because of all the astroturfing or uninformed voices but I think this is just a case of perverse incentives where miners do not have the best interests for bitcoin at heart because I have not seen any academic say increasing the blocksize is not risky while I have seen many that have said it is risky.  The rest is just noise and distraction originating in those who benefit from a held back bitcoin.

[achow101]

2.  Even if UASF is supported by the community, if LN isn't ready I'm afraid from a practical point of view users will turn to a BU fork because of lower transaction costs even if they themselves think BU is a shit long term scaling solution. If you needed to send a $10 payment and had some money on an exchange, wouldn't you buy the BU coins with your BTC and initiate a $0 fee transfer instead of paying $5 in fees?

Segwit in itself provides a capacity increase that will, in theory, decrease transaction fees like the BU fork would, in theory. It does not exist to just enable LN but is itself a way to increase the number of transactions that can fit in a block.

3.  If LN isn't instantly deployable how long will it take to be deployed?  Will sites like exchanges which create an enormous amount of on chain transfers be able to instantly reduce their transfer load through LN and if so by how much?

LN is still being worked on but it is available on testnet, so it certainly is close to being deployed. I don't think it is instantly deployable though. I don't think it would be particularly useful for exchanges as LN is (so far) only useful for small, recurring payments. Most exchanges don't perform on chain transactions except when their customers are depositing or withdrawing money. The exchanging itself is all internal database records.

4.  A continuation of question 3, in terms of development time and effort what cost will merchants pay to switch to using LN?  I'm not saying what it will take them to recognize non-segwit transactions on the segwit chain, I know that is trivial.  I am asking how long will it take and how tedious will it be for them and their users to start benefiting from lower transaction costs.

Given that there will be software available that can be used by other customers and merchants, it shouldn't really be that much effort to switch to using LN. The most work will be from setting up payment channels, but in most scenarios, the merchant would only be receiving money, not sending, so they risk almost no money. The only issue there is that they would not be able to access the money until after the channel is closed.

5.  If we agree that the ASICBOOST miners have acted Machiavellian up to this point, why exactly would we ever expect them to honor any agreement where they would sacrifice their profitability?  If they lose the ability to exploit ASICBOOST, that means they could go from being high profit miners to being losers in the mining space no matter what price bitcoin reached. If they no longer have a competitive advantage against other miners, they will lose money regardless of the BTC price.  With this in mind how is a malicious attack on BIP148 different from their secretive ASICBOOST exploitation and the faux politics of certain miners?  Why would they possibly not attack the BIP148 chain if what they have done so far is actually worse?  A long chain reorg that causes user losses would be worse but an above board attack where BIP148 users are not given a false sense of security would not be worse, it would just be a continuation of disruption.

Not using ASICBOOST is still profitable. Just because ASICBOOST miners stop using ASICBOOST does not mean that they will suddenly be mining at a loss. It just means that their profits will decrease.

8.  LTC already has working Segwit but doesn't afaik have LN set up on it.  Why is this?  When will LN be deployable on LTC?

Because all LN development has been for Bitcoin. While Bitcoin and litecoin are similar, there is still a non-trivial amount of work to do to make LN work on Litecoin. Fron what I understand, those working on LN will finish it for Bitcoin first (except for segwit) and then port it over to Litecoin once the Bitcoin version is completed.

9.  How could a post-UASF chain ever truly know within a reasonable time period that there will not be a reorganization?  If we assume that there isn't any decrease in the ASICBOOST hashing power and that a large amount of malicious miners have not been dormant in anticipation and or are not just coming online can we be confident in the BIP148 chain?  Are those reasonable assumptions to make?

A BIP148 chain cannot be reorged with a non-BIP 148 chain. This is because it is a restriction of the current consensus rules so BIP 148 nodes considers the non-BIP 148 chain to be invalid. However the non-BIP 148 chain can be reorged with the BIP 148 chain since non-BIP 148 nodes still consider the BIP 148 chain to be valid.

10.  To end off on a note of more personal politics, I think both Core and BU have a terrible actor or two.  Ver is obviously not honest and Theymos who I am aware doesn't work for Blockstream could have done a much better job at constructively moderating /r/bitcoin.

Theymos is not involved with Core at all. He has never contributed to Bitcoin Core and does not participate in its development whatsoever.

[cahulahupo]

Thank you for your response but most answers you've given failed to answer the questions.  Perhaps I wasn't always clear enough.  Some also show your ignorance though.  I really hope I can get some answers by real developers who understand the issue.

2.  Segwit does provide a capacity increase but as I understand it the capacity increase only happens with actual segwit type transactions are being used.  If they will not be instantly used because of a fear of a reorg then transfer fees will remain high possibly while BU will immediately have low transfer fees.  And yes you are mistaken. A reorg is possible with BIP148 if there is a malicious attack. A reorg going back to Aug 1 could happen in September for all we know.

3.  I know the difference between an on chain transfer and an off chain transfer.  I assume most people send and receive BTC from their exchange wallets though.  Maybe that is a bad assumption.  I assume regular users will not be making use of LN right away so are you saying that the effects of a post segwit BTC might not be felt for months?  BU might be near free to exchange for months while BTC continues to have $5 transfer fees?  Is that possible?

5.  You say that not using ASICBOOST is still profitable but there is absolutely no way that you can know this.  Even if it is profitable a miner's ROI may go down from 30% to 3% meaning that BTC would have to explode to $30,000 for the miner to benefit as much as they benefited with ASICBOOST.  If a miner's ROI is 5% and goes down to 0.05% BTC would need to be $300,000.  And of course there is the possibility that those abusing ASICBOOST would in fact lose money if they were to honestly mine.

9.  You completely miss the point.  A BIP148 chain can be reorged with a BIP148 chain that ASICBOOST miners have been honestly but covertly mining.  They may wait a month to release a completely reorged month long chain.  Segwit Bitcoin would essentially be dead.  This is a real threat.

10.  I never said Theymos was involved with Core but his actions are the faux justification for ASICBOOST abusers and Ver's temper tantrum.

[thejaytiesto]

2.  Even if UASF is supported by the community, if LN isn't ready I'm afraid from a practical point of view users will turn to a BU fork because of lower transaction costs even if they themselves think BU is a shit long term scaling solution. If you needed to send a $10 payment and had some money on an exchange, wouldn't you buy the BU coins with your BTC and initiate a $0 fee transfer instead of paying $5 in fees?

Why not just buy Litecoin and use that to transact? It's pretty stupid to trust your money with BU when the software has crashed so many times. LTC on the other hand is a BTC Core clone with segwit enabled, so I can't think of a better alternative right now in case things go south in august 1st.

In the case a 2-coin scenario... it's stupid to sell any side. No one knows what's going to happen. Im holding both so good luck with those that actually dump for whatever chain they think will survive.

[mocacinno]

LN is running on testnet4 of litecoin... I did a testsetup a couple weeks ago... Got it running (more or less). Just for fun, see if it would do something.
Ran https://github.com/ltcsuite/ltcd on the testnet, then ran https://github.com/lightningnetwork/lnd
 

[achow101]

Thank you for your response but most answers you've given failed to answer the questions.  Perhaps I wasn't always clear enough.  Some also show your ignorance though.  I really hope I can get some answers by real developers who understand the issue.

I am a real developer. I am a contributor to Bitcoin Core.

2.  Segwit does provide a capacity increase but as I understand it the capacity increase only happens with actual segwit type transactions are being used.  If they will not be instantly used because of a fear of a reorg then transfer fees will remain high possibly while BU will immediately have low transfer fees. 

Why would it not be used because of a fear of a reorg? Segwit will only activate if 95% of the hashrate signals that they are enforcing it. Unless the majority of those signalling are false signalling, there won't be a reorg.

And yes you are mistaken. A reorg is possible with BIP148 if there is a malicious attack. A reorg going back to Aug 1 could happen in September for all we know.

How so? A non BIP 148 chain is invalid to a BIP 148 chain so there is no reorg possibility there. The only reorg risk is if a valid but different BIP 148 chain were mined. That would cause a reorg, but that is a risk for all blockchains, not just BIP 148.

3.  I know the difference between an on chain transfer and an off chain transfer.  I assume most people send and receive BTC from their exchange wallets though.  Maybe that is a bad assumption.  I assume regular users will not be making use of LN right away so are you saying that the effects of a post segwit BTC might not be felt for months?  BU might be near free to exchange for months while BTC continues to have $5 transfer fees?  Is that possible?

That is certainly possible, but again, you are assuming that people won't be using segwit. As I said earlier, segwit does not exist to just enable LN. Furthermore, even though we may not see the capacity increases of segwit immediately, users will still benefit from lower fees from using segwit. The fees that a segwit transaction pays is smaller than the fee of the non-segwit version of that transaction. There is an immediate benefit to using segwit as the amount you pay in fees will be lower, even though the fee rate may not decrease immediately.

5.  You say that not using ASICBOOST is still profitable but there is absolutely no way that you can know this.  Even if it is profitable a miner's ROI may go down from 30% to 3% meaning that BTC would have to explode to $30,000 for the miner to benefit as much as they benefited with ASICBOOST.  If a miner's ROI is 5% and goes down to 0.05% BTC would need to be $300,000.  And of course there is the possibility that those abusing ASICBOOST would in fact lose money if they were to honestly mine.

The majority of the network is not using ASICBOOST simply because you need a pool that can support it, and no pool currently supports ASICBOOST. This indicates that mining without ASICBOOST is still profitable.

9.  You completely miss the point.  A BIP148 chain can be reorged with a BIP148 chain that ASICBOOST miners have been honestly but covertly mining.  They may wait a month to release a completely reorged month long chain.  Segwit Bitcoin would essentially be dead.  This is a real threat.

As I said earlier in this post, that is a risk that all blockchains have. However that means that those miners would be diverting hash power away from a chain that they think is more likely to earn them money. By covertly mining a reorg chain for BIP 148 nodes, those miners are wasting their electricity on something that they think won't earn any money (and probably won't if there are reorgs).

[DannyHamilton]

1.  How is it possible that any pro-segwit movement can reach any significant amount of mining power support because in the mining world of perfect competition only those taking advantage of covert ASICBOOST should be able to survive?  Shouldn't all honest miners be losing money in the long run, at least if difficulty adjusted faster to price increases?  Perhaps by August 1st the ASICBOOST miners will have been limited by development time frames but surely if the UASF is not successful by Aug 1st, bitcoin will never be able to adopt segwit with any realistic amount of mining power.  That is because ACISBOOST abusers will not only be self-sufficient but much more flush with cash to reinvest in more hashing power until honest miners are extinct from uncompetitiveness.  Am I correct in my assumptions?

There's a lot of prejudice and judgement in that question.

There is nothing "dishonest" or "abusive" about ASICBOOST.  ASICBOOST is a performance optimization.

That being said, there are additional variables that come into play when determining which miners maintain profitability and which don't survive.  Some of the other factors that influence that result are local cost of electricity, best bitcoin to local currency exchange rate they can find, and strength of the local currency against foreign currencies. Additionally, there are logistics to be managed in acquiring, setting up, and maintaining the mining hardware.

Furthermore, it would be possible to implement SegWit in the future as a hard fork if necessary.  If that were to happen, then ASICBOOST would be unable to attack the resulting blockchain and as long as enough nodes and merchants adopted the new rules ASICBOOST miners would lose their competitive advantage.

It is also possible that in the future someone else comes up with another mining optimization that is better than ASICBOOST and is compatible with a SegWit soft fork.

2.  Even if UASF is supported by the community, if LN isn't ready I'm afraid from a practical point of view users will turn to a BU fork because of lower transaction costs even if they themselves think BU is a shit long term scaling solution. If you needed to send a $10 payment and had some money on an exchange, wouldn't you buy the BU coins with your BTC and initiate a $0 fee transfer instead of paying $5 in fees?

That might be possible, but there is no indication that BU is going to fork any time soon.

5.  If we agree that the ASICBOOST miners have acted Machiavellian up to this point,

And if we don't agree on that?

why exactly would we ever expect them to honor any agreement where they would sacrifice their profitability?  If they lose the ability to exploit ASICBOOST, that means they could go from being high profit miners to being losers in the mining space no matter what price bitcoin reached. If they no longer have a competitive advantage against other miners, they will lose money regardless of the BTC price.

If those miners feel that they'd be better off with a smaller percentage of a larger value than a larger percentage of a smaller value, then they will be motivated.

Would you rather have 90% of $1, or 1% of $1,000,000?