[bitcoin-dev] Why Satoshi's temporary anti-spam measure isn'ttemporary

Thu Jul 30 13:33:09 UTC 2015

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Jorge,

You know, it is always insightful to get the perspective of active
participants and Core developers like yourself. As Adam pointed out
earlier, the developers have done mileage in this space and have
already considered most of the conceptual issues and technical
challenges that must resurface in waves as new interested parties join
the list. Allow me, in this response to your message, to make a
proposal to those who may be interested:

Bitcoin's protocol functions and the implications of this innovation
for the future are difficult to grasp, even for the smartest among us.
Then there are also the words of Niels Bohr:

"Prediction is difficult, especially about the future."

They say a lot of time and energy is wasted because we don't know what
we don't know. Years of discussion among those in the list has
established certain axioms that determine the options for Bitcoin
going forward. According to my comprehension, the following are some
of the most relevant for the present discussion (please correct me
where I'm off the mark):

1. A high degree of decentralization is prima optima.

2.  Bitcoin is much more than a payment network. A lot of the
non-payment features are, arguably, what gives Bitcoin most of its
value. Yet, the payment functionality is a major design feature and
all agree that it should scale - subject to axiom 1.

3. The Bitcoin payment network ("Layer 1"), due to technical
constraints imposed by its p2p design, cannot compete with Visa and
other centralized transmission channels for speed or transaction
volume. Nor can it handle the transaction requirements of the world's
population - the scaling required would necessarily render Bitcoin
centralized, insecure and, therefore, worthless.

4. The addition of "layer 2" protocols (such as Lightning and other
sidechains) will allow fast, low-fee (and with virtually instant
confirmation) bitcoin transactions within two years, according to the
developers active in that:
http://www.youtube.com/watch?v=jE_elgnIw3M
http://www.youtube.com/watch?v=fBS_ieDwQ9k

5. This "layering of protocols" simplifies the scaling (blocksize)
debate because it separates
 A) the primary concern for security and fidelity via
decentralization, and
 B) the ideal of universal accessibility via fast, low-fee transactions.
Discussion about scalability can therefore proceed with the knowledge
that Lightning and other "layer 2" sidechains will make Bitcoin
accessible to the global majority - and be fast like Bruce Lee - while
the Bitcoin developers can focus on making Bitcoin Core protocol
(layer 1) the world heavyweight champion - Muhammad Ali.

Since I've maintained your interest up to the final sentence, I say:
as an insurance against a capacity crisis before layer 2 is deployed,
why not implement bip100's 2MB blocksize proposals in a testnet?

On 07/30/2015 04:38 PM, Jorge Timón wrote:
> It is important ro note that even if lightning was never developed,
> the block size remains at 1 MB forever and fees rise to 10 usd per 
> transaction, such "high fees" are still extremely competitive with 
> non-decentralized payment systems that have proportional fees. For 
> example, 10 usd is still lower than 1% when you are moving more
> than 1000 usd. I know, this doesn't work for micro-transactions,
> but I don't think Bitcoin can be useful for micro-transactions in
> the long term unless something like lightning payment channels is
> deployed. Until we accept the second fact, it will be very hard to
> discuss any projection of future usage. I think that believing that
> all the transactions of the entire world population can be made
> in-chain while keeping bitcoin decentralized is incredibly naive.
> Not even nasdaq has that capacity (and if full node's require
> nasdaq's capacity, I don't think we can talk about a decentralized
> system anymore).
> 
> On Jul 30, 2015 11:16 AM, "Venzen Khaosan via bitcoin-dev" 
> <bitcoin-dev at lists.linuxfoundation.org 
> <mailto:bitcoin-dev at lists.linuxfoundation.org>> wrote:
> 
> Adam,
> 
> - From your explanation it is evident that fast, cheap bitcoin 
> transactions are possible. It is encouraging that Bitcoin _can_
> indeed compete with Visa, Paypal, et al. via Layer 2 protocols such
> as Lightning.
> 
> The youtube interview with you and Greg re: Lightning requires
> some concentration and I'll have to watch it another couple of
> times to better grasp everything that is explained about the
> protocol and its interaction with Bitcoin.
> 
> Thank you for your considered and informative response, else
> Raystonn and I might have gotten in an unnecessary scrap about
> fees, economics and what not.
> 
> regards, Venzen Khaosan
> 
> 
> 
> On 07/30/2015 10:49 AM, Adam Back wrote:
>> I dont think people consider other blockchains as a competitive 
>> threat.  A PoW-blockchain is a largely singleton data structure 
>> for security reasons (single highest hashrate), it is hard for
>> an alternative chain to bootstrap or provide meaningful
>> security. Secondly the world largely lacks expertise to maintain
>> a blockchain to bitcoin's security level, perhaps you can see a
>> hint of this in the recently disclosed security vulnerability by
>> Pieter Wuille and Gregory Maxwell.  Calls to this as an argument
>> are not resonating and probably not helping your argument.
>> Bitcoin has security properties, and a competing system cant
>> achieve better properties by bypassing security, any blockchain
>> faces the same fundamental security / decentralisation
>> limitations.
> 
>> Secondly Bitcoin can obviously compete with itself with
>> different parameters and defacto *does* today.  I think it is a
>> safe estimate that > 99% of Bitcoin transactions right now are
>> happening in Bitcoin related systems with various degrees of
>> audit, reconciliation, provable reserves etc.  I think we can
>> expect this to continue and become more secure via more
>> reconciliation, and longer term via lightning or Bitcoin
>> sidechains with different parameters.  It is a different story to
>> have a single central system (Bitcoin with parameters changed to
>> the point of centralisation failure) vs having multiple choices,
>> because some transactions can more easily use relatively
>> centralised systems (eg micropayments), and more interestingly
>> the combination of a secure and decentralised layer 1 plus
>> choices of less decentralised layer 2 options, can be interesting
>> because the layer 2 is provided cover from attack.  There is less
>> to be gained by attacking relatively centralised layer 2 because
>> any payments at risk of policy abuse (which is typically a small
>> subset) can easily switch to layer 1. That in itself makes layer
>> 2 transactions also less susceptible to policy abuse.  Further
>> lightning it appears from work so far should add significant
>> scale while retaining trustlessness and a good degree of
>> decentralisation.
> 
>> Finally you seem to be focusing on "artificial" limits where
>> that is not the issue under consideration.  The limits are
>> technical and relating to decentralisation and security.  I wont
>> go over them again as this topic has been covered many times in
>> recent months. Any chain that tried to go to extreme parameters
>> (very low block intervals, or very large blocksizes) would have
>> the same decentralisation problems as Bitcoin would if it did the
>> same thing.  There are a number of alt coins that have failed as
>> a result of poor parameter choices, there are inherent security 
>> limits.
> 
>> Adam
> 
>> ps Etiquette note for yourself and others: please dont be 
>> repetitive or attempt to be forceful.  Many people have spent
>> many years understanding this very complex system, from my own 
>> experience it is rare indeed to think of an entirely new concept
>> or analysis, that hasnt' been long considered and put to bed 3 or
>> 4 years ago. Thoughtful polite and constructive comments are
>> welcome but I recommend to not start from an assumption that you
>> have a clear and better insight than the entire technical
>> community, because I have to say from my own experience that is
>> very rarely the case.  It can be useful to test theories on
>> #bitcoin IRC channel to find out what has been already concluded,
>> find the references and avoid having to have that hashed out on
>> this list which is trying to be focussed on technical solutions.
> 
> 
>> On 29 July 2015 at 16:10, Raystonn . via bitcoin-dev 
>> <bitcoin-dev at lists.linuxfoundation.org
> <mailto:bitcoin-dev at lists.linuxfoundation.org>> wrote:
>>>> Cheapest way to send value? Is this what Bitcoin is trying
>>>> to do? So all of the smart contract, programmable money,
>>>> consensus coding and tremendous developer effort is bent to
>>>> the consumer demand for cheaper fees. Surely thou jests!
>>> 
>>> 
>>> These other features can be replicated into any alternative 
>>> blockchain, including those with lower fees.  In the
>>> open-source world of cryptocurrency, no feature will remain a
>>> value-add for very long after it has been identified to be
>>> such.  Anything adding value will quickly be absorbed into
>>> competing alternative blockchains.  That will leave economic
>>> policy as the distinguishing factor.
>>> 
>>>> ... it is not the case ... that reluctance to concede
>>>> blocksize is an attempt to constrain capacity. Greg Maxwell
>>>> thoroughly explained in this thread that the protocol's
>>>> current state of development relies on  blocksize for
>>>> security and, ultimately, as a means of protecting its degree
>>>> of decentralization.
>>> 
>>> 
>>> A slow or lack of increase to maximum transaction rate will
>>> cause pressure on fees.  Whether this is the desired goal is
>>> not relevant.  Everyone has agreed this will be the outcome.
>>> As to a smaller block size being needed for additional
>>> decentralization, one must simply ask how much we are all
>>> willing to pay for that additional decentralization.  It is
>>> likely that the benefit thereto will have to be demonstrated by
>>> some power attacking and destroying a less decentralized
>>> currency before the benefit of this feature is given monetary
>>> value by the market.  Until then, value will bleed to the
>>> network with the least friction, because it will have the
>>> greatest ability to grow its network effect. That means the
>>> blockchain with adequate features and cheapest fees will
>>> eventually have the largest market share.
>>> 
>>> 
>>> -----Original Message----- From: Venzen Khaosan Sent:
>>> Wednesday, July 29, 2015 3:11 PM To: Raystonn . Cc: 
>>> bitcoin-dev at lists.linuxfoundation.org
> <mailto:bitcoin-dev at lists.linuxfoundation.org> Subject: Re: 
> [bitcoin-dev]
>>> Why Satoshi's temporary anti-spam measure isn'ttemporary
>>> 
>> Raystonn, I'm aware that you're addressing your question to Greg 
>> Maxwell, however a point you keep stating as fact calls for 
>> reference:
> 
>> On 07/30/2015 04:28 AM, Raystonn . via bitcoin-dev wrote: [snip]
>>>>> 
>>>>> How do you plan to address the bleeding of value from
>>>>> Bitcoin to alternative lower-fee blockchains created by
>>>>> the artificially-high bitcoin transaction fees when users
>>>>> begin looking for the cheapest way to send value?
> 
>> Cheapest way to send value? Is this what Bitcoin is trying to
>> do? So all of the smart contract, programmable money, consensus
>> coding and tremendous developer effort is bent to the consumer
>> demand for cheaper fees. Surely thou jests!
> 
>>>>> Modern economic study has shown that liquidity moves to
>>>>> the location of least friction.
> 
>> Modern economic study? Can you please provide a link or
>> reference to the study you are referring to.
> 
>> "liquidity moves to the location of least friction"
> 
>> This sounds like "econo-speak" and makes no sense. The
>> definition of Liquidity is the degree to which an asset/security
>> can be bought or sold in the market without affecting the price.
> 
>> That is why bitcoin is said to have low liquidity: buying or 
>> selling only 100 BTC visibly affects the exchange price. You 
>> probably mean "people like cheap fees", which is true, but as 
>> others have said, because of Bitcoin's powerful features, they
>> are willing to pay higher fees and wait longer for transactions
>> to execute.
> 
>> As for your public cross-examination of Greg Maxwell, your case 
>> seems to  be made on the assumption that limiting the size of
>> the blockchain is an attempt to artificially raise tx fees, but
>> it is not the case (as you and others repeatedly argue) that
>> reluctance to concede blocksize is an attempt to constrain
>> capacity. Greg Maxwell thoroughly explained in this thread that
>> the protocol's current state of development relies on  blocksize
>> for security and, ultimately, as a means of protecting its degree
>> of decentralization.
> 
>> Surely, this is an obvious concern even for those who are 
>> campaigning for the hare-brained ideal of making Bitcoin a
>> "faster, cheaper alternative" to visa or paypal? If we lose 
>> decentralization, we lose the whole thing, right? Incorrect or 
>> correct?
>>> _______________________________________________ bitcoin-dev 
>>> mailing list bitcoin-dev at lists.linuxfoundation.org
> <mailto:bitcoin-dev at lists.linuxfoundation.org>
>>> https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev
> _______________________________________________ bitcoin-dev mailing
> list bitcoin-dev at lists.linuxfoundation.org 
> <mailto:bitcoin-dev at lists.linuxfoundation.org> 
> https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev
> 
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1

iQEcBAEBAgAGBQJVuieTAAoJEGwAhlQc8H1meAIH/RHUV72bbMItZOT7rvhEU56r
lqEcvwXBSCYgsh1ieVeTdC/ydJnRSzWsdZxM3D7PEOzutlG+VaJQVSJREItGb2GW
PYiZ3uwSwF64nRq5bZ7aS2pT/Zo1a1yAf4H5rbeyxxoWC+zkmSsmcf73MgmslIuU
7XXHNztCX3glfOctr+J61WEKBw0ItQCTsp9J08yVlj/gvKTi3U2jDcYV5mf/3D0j
pvXl244DG4b+nYetRyyonYbZelSUYfCghNBJhUYZApVmcgfKDRPeX1uWfkl0HuUd
Kc+uZtrhJaUXdRlqc50nOsRSCAK+d4PGClF8JFlzI65+SG7VzkVqc8SkSDfNXfY=
=r4SA
-----END PGP SIGNATURE-----