[bitcoin-dev] New difficulty algorithm part 2

ZmnSCPxj ZmnSCPxj at protonmail.com
Fri Oct 13 04:45:33 UTC 2017

Good morning,

>ZmnSCPxj wrote:
>> Thus even if the unwanted chain provides 2 tokens as fee per block,
>> whereas the wanted chain provides 1 token as fee per block, if the
>> unwanted chain tokens are valued at 1/4 the wanted chain tokens, miners
>> will still prefer the wanted chain regardless.
>This is a good point I was not thinking about, but your math assumes
>1/2 price for a coin that can do 2x more transactions. Holders like
>Roger Ver have an interest in low price and more transactions. A coin
>with 2x more transactions, 22% lower price, and 22% lower fees per
>coin transferred will attract more merchants, customers, and miners
>(they get 50% more total fees) and this will in turn attract more
>hodlers and devs. This assumes it outweighs hodler security concerns.
>Merchants and customers, to the extent they are not long term hodlers,
>are not interested in price as much as stability, so they are somewhat
>at odds with hodlers.

As of this moment, BT1 / BT2 price ratio in BitFinex is slightly higher than 7 : 1.  Twice the transaction rate cannot overcome this price ratio difference.  Even if you were to claim that the BitFinex data is off by a factor of 3, twice the transaction rate still cannot overcome the price ratio difference.  Do you have stronger data than what is available on BitFinex?  If not, your assumptions are incorrect and all conclusions suspect.

>Bitcoin consensus truth is based on "might is right". Buyers and
>sellers of goods and services ("users") can shift some might to miners
>via fees, to the chagrin of hodlers who have more interest in security
>and price increases. Some hodlers think meeting user needs is the
>source of long term value. Others think mining infrastructure is.

Mining infrastructure follows price.  If bitcoins were still trading at 1 USD per coin, nobody will build mining infrastructure to the same level as today, with 5000 USD per coin.

Price will follow user needs, i.e. demand.

>seem to require hodlers to correctly identify and rely solely on good

For the very specific case of 2X, it is very easy to make this identification.  Even without understanding the work being done, one can reasonably say that it is far more likely that a loose group of 100 or more developers will contain a few good or excellent developers, than a group of a few developers containing a similar number of good or excellent developers.

User needs will get met only on the chain that good developers work on.  Bitcoin today has too many limitations: viruses on Windows can steal all your money, fee estimates consistently overestimate, fees rise during spamming attacks, easy to lose psuedonymity, tiny UTXOs are infeasible to spend, cannot support dozens of thousands of transactions per second.  Rationally, long-term hodlers will select a chain with better developers who are more likely to discover or innovate methods to reduce, eliminate, or sidestep those limitations.  Perhaps the balance will change in the future, but it is certainly not the balance now, and thus any difficulty algorithm change in response to the current situation will be premature, and far more likely to cause disaster than avert one.

>Whatever combination of these is the case, bad money can
>still drive out good, especially if the market determination is not
>A faster measurement of hashrate for difficulty enables the economic
>determination to be more efficient and correct. It prevents the
>biggest coin from bullying forks that have better ideas. Conversely,
>it prevents miners from switching to an inferior coin simply because
>it provides them with more "protection money" from fees that enables
>them to bully Bitcoin Core out of existence, even in the presence of a
>slightly larger hodler support.

This requires that all chains follow the same difficulty adjustment: after all, it is also entirely the possibility that 2X will be the lower-hashrate coin in a few months, with the Core chain bullying them out of existence.  Perhaps you should cross-post your analysis to bitcoin-segwit2x also.  After all, the 2X developers should also want to have faster price discovery of the true price of 2X, away from the unfavorable (incorrect?) pricing on BitFinex.

>Devs are a governing authority under the influence of users, hodlers,
>and miners. Miners are like banks lobbying government for higher total
>fees. Hodlers are the new 1%, holding 90% of the coin, lobbying both
>devs and users for security, but equally interested in price
>increases. Users are "the people" that devs need to protect against
>both hodlers and miners. They do not care about price as long as it is
>stable. They do not want to become the 99% owning 10% of the coin or
>have to pay unecessary fees merely for their coin to be the biggest
>bully on the block. A faster responding difficulty will take a lot of
>hot air out of the bully. It prevents miners from being able to
>dictate that only coins with high fees are allowed. They are less
>able to destroy small coins that have a fast defense.
>The 1% and banks would starve the people that feed them to death if
>they were allowed complete control of the government. Are hodlers and
>miners any wiser?

Are developers any wiser, either?

Then consider this wisdom: The fewer back-incompatible changes to a coin, the better.  Hardforks of any kind are an invitation to disaster and, at this point, require massive coordination effort which cannot be feasibly done within a month.  Fast market determination can be done using off-chain methods (such as on-exchange trades), and are generally robust against temporary problems on-chain, although admittedly there is a counterparty risk involved.  The coin works, and in general there is usually very little need to fix it, especially using dangerous hardforks.

>Devs need to strive for an expansion of the coin
>quantity to keep value constant which is the foundation of the 5
>characteristics of an ideal currency.

Is that your goal?  This is a massive departure from the conception of Bitcoin as having a fixed limit and effectively becoming deflationary.  It will also lead to massive economic distortions in favor of those who receive newly-minted coins.  I doubt any developer would want to have this property.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linuxfoundation.org/pipermail/bitcoin-dev/attachments/20171013/1afdebf0/attachment-0001.html>

More information about the bitcoin-dev mailing list