Bitcoin 20MB Fork

[amincd]

Hey I'd be fine with a 5 MB block size limit, and some percentage increase every year afterward. I think 20 MB would probably be better, but that's far less important than putting in place some dynamic hard limit, ANY dynamic hard limit, that steadily increases, year by year, so that we never have to do a hard fork again to change the limit.

Should we need a stricter limit in the future, it should be done through a SOFT FORK, with >50% of miners enforcing it. Static hard limits are dangerous, and should not be the way Bitcoin controls bloat.

[1715138948]

1715138948

[1715138948]

1715138948

[1715138948]

1715138948

[1715138948]

1715138948

[1715138948]

1715138948

[hdbuck]

[iCEBREAKER]

Hey I'd be fine with a 5 MB block size limit, and some percentage increase every year afterward. I think 20 MB would probably be better, but that's far less important than putting in place some dynamic hard limit, ANY dynamic hard limit, that steadily increases, year by year, so that we never have to do a hard fork again to change the limit.

Should we need a stricter limit in the future, it should be done through a SOFT FORK, with >50% of miners enforcing it. Static hard limits are dangerous, and should not be the way Bitcoin controls bloat.

I agree with the goal of never again having to do a hard fork to change the limit, but am not sure if linear increases are appropriate for a system that could grow geometrically.

Could you support starting with a 2MB cap that then doubles every year?

I'd might be OK with a 5MB cap that doubles when the block reward halves, depending on how it effects TOR/DSL users.

[solex]

I agree with the goal of never again having to do a hard fork to change the limit, but am not sure if linear increases are appropriate for a system that could grow geometrically.

Could you support starting with a 2MB cap that then doubles every year?

I'd might be OK with a 5MB cap that doubles when the block reward halves, depending on how it effects TOR/DSL users.

Between 2MB and 5MB is a reasonable starting point, although the 2MB is too tight for very long. It doesn't even have to double every year, just every two years to match bandwidth improvement:

http://www.nngroup.com/articles/law-of-bandwidth/

Summary: Users' bandwidth grows by 50% per year (10% less than Moore's Law for computer speed). The new law fits data from 1983 to 2014.

Nielsen's Law of Internet bandwidth states that:

a high-end user's connection speed grows by 50% per year
The dots in the diagram show the various speeds with which I have connected to the Net, from an early acoustic 300 bps modem in 1984 to an ISDN line when I first wrote this article (and updated to show the 120 Mbps upgrade I got in 2014). It is amazing how closely the empirical data fits the exponential growth curve for the 50% annualized growth stated by Nielsen's Law. (The y-axis has a logarithmic scale: thus, a straight line in the diagram represents exponential growth by a constant percentage every year).

[CoinCidental]

"TOR" is shorthand for low-bandwidth connections, whether privacy-enhanced darkweb internets-within-the-internet or plain old slow DSL.

You need to keep up.  This has already addressed here:

I view this as paranoia. While you and tvbcof are using lots of services outside of TOR you don't want to use Bitcoin outside of TOR. Why don't you try to impose TOR on all websites if we are at this subject? Only limit only to Bitcoin? Why should the network comply with your view and fears? Why don't you let the network decide this as it should happen?

Concern for low-bandwidth DSL and hardened TOR-like connections wasn't limited to tvbcof and myself.

It was also voiced by TF and Theymos.  Why did you edit that part out?

Here it is again, with red bolding to help you learn.  Perhaps this time you'll be able to read, and maybe even respond, with on-point arguments rather than personal attacks like accusations of "paranoia."

Exactly what block size the network can support is very much debatable. I currently think that 10 MB would be fine and 50 MB would be too much, though these are mostly just feelings. There should be more rigorous study of the actual limits of the network. (Gavin's done some nice work on the software/hardware front, though I'm still worried about the capabilities of typical Internet connections, and especially how they'll increase over time.)

Exactly. 30 kBps upload is common in Australia, and you sure should be able to run a full node in a typical internet connection in Australia, or Brazil, or Philippines, or whatever. The block size needs to be useful for the (lowest reasonable) common denominator, not the median.

IMO 10 MB is too much, maybe 5 MB.

youre missing the whole argument ,every bitcoin user in brazil ,phiilipines or australia doesnt have to run a full node
thats what lightweight clients are for

even a 10 year old computer on dial up could  run multibit or a lite wallet

i visit a lot of countries  around the world and some of them have suprisingly fast internet

http://www.netindex.com/download/allcountries/

[davout]

And it's only 3 transactions per second, not 7.

Ok lol.

[Stephen Gornick]

Without a fork, does tumbling become prohibitively expensive?

[traincarswreck]

Some of the not so accepted science, but yet common sense stuff, suggests much of the proposals will suffice because in the future some of these "parameters" may not continue to (or necessarily must continue to) increase. In a more cooperative/intelligent society the need for security doesn't seem to me like it will increase at the same rate.  The "direct" block-chain" based transactions don't seem like they need to continue either, especially geometrically.  This is stuff can't be accurately forecast, but it should help resolve many to a middle way type solution.

Also I read about this guy who declares war on such a 'fork' process, it might be nice to kill two birds with one stone and show that bitcoin is capable of handling both issues.  It reminds me of the 140 so years of creation of coins, that I suspect is fairly arbitrary to some degree. As long as there is a psychological stability to the supply any amount of years beyond say 20 would probably be fine.  

This is the blood and guts of AI right? The same person that points that bitcoin is not just a money probably will call me a quack for pointing out exactly what we all know: trustworthy computing is the foundation for it.

And so in the future what is a "transaction", and what is the role of "money"?  If we had AI, what would be the importance of block size?  Its just, clearly if we understand trustworthy computing, smart contracts, the coming evolution of games...we should know that AI is coming within a few years and so much like thinking of what is "Ideal Money" can help us extrapolate a solution towards it (bitcoin), think about the truth of what the internet will become, will help us understand the direction to take bitcoin with.

I suspect now, this problem we are discussing is the last mechanism before global currency stability and then "price" movement, and that many peoples in the community are quietly frustrated because we need consensus and finality on the issue.

I say this...admittedly, walking into this dialogue from a few pages back, but only because a relevant article was quoted (and because seemingly nobody understands "ideal money")

[amincd]

I agree with the goal of never again having to do a hard fork to change the limit, but am not sure if linear increases are appropriate for a system that could grow geometrically.

Could you support starting with a 2MB cap that then doubles every year?

I'd might be OK with a 5MB cap that doubles when the block reward halves, depending on how it effects TOR/DSL users.

Yes, a geometric increase is what I meant. Gavin's current proposal is for it to double every two years, meaning it increases 1.4X every year, which gives it a reasonably high chance of staying behind consumer bandwidth growth (and thus keeping the network highly decentralized).

So to answer your question, yes I would support something along the lines of what you're proposing. I think 5 MB would be much better than 2 MB though, because Bitcoin has a tendency to see sudden spurts in adoption, and so it'd be nice for it to have some room to grow quickly in the coming years.

http://www.nngroup.com/articles/law-of-bandwidth/

Summary: Users' bandwidth grows by 50% per year (10% less than Moore's Law for computer speed). The new law fits data from 1983 to 2014.

Nielsen's Law of Internet bandwidth states that:

a high-end user's connection speed grows by 50% per year
The dots in the diagram show the various speeds with which I have connected to the Net, from an early acoustic 300 bps modem in 1984 to an ISDN line when I first wrote this article (and updated to show the 120 Mbps upgrade I got in 2014). It is amazing how closely the empirical data fits the exponential growth curve for the 50% annualized growth stated by Nielsen's Law. (The y-axis has a logarithmic scale: thus, a straight line in the diagram represents exponential growth by a constant percentage every year).

Thanks for posting.

And I think one thing that often gets lost in the discussion on the hard limit, and that I can't stress enough:

There are other ways to stop bloat besides a hard limit in the protocol. A protocol limit is the most crude and inflexible way to counter spam/bloat. If a 40-50% increase per year ends up being faster than connection speed increases, it is very unlikely that other means will not be found to limit block sizes.

[traincarswreck]

"How do 'good money' and 'bad money' differ, if at all, for the valuable function of facilitating utility transfer?" ... if we consider contracts having a relatively long time axis then the difference can be seen clearly.

   Consider  a society where the money in use is subject to a rapid and unpredictable rate of inflation so that money worth 100 now might be worth from 50 to 10 by a year from now.  Who would want to lend money for the term of a year? In this context we can see how the "quality" of a money standard can strongly influence areas of the economy involving financing with longer-term credits.

   And also, if we view money as of importance in connection with transfers of utility, we can see that money itself is a sort of "utility", using the world in another sense, comparable to supplies of water, electric energy or telecommunications.  And then, if we think about it, we can consider the quality of money as comparable to the quality of some "public utility" like the supply of electric energy or of water.

What seems interesting, aside from the obsession with the significance of money in relation to contracts, is that seemingly we are finally stating to understand money in relation to telecommunications like this very old lecture.

You see, I get accused of things such as only thinking of bitcoin as a money.  But I think largely as a whole we have a giant block to the obvious reality of the mechanisms we are creating "socially":

This memorandum concerns some ideas for new designs of the control system in high-speed digital computers. The ideas are yet in an immature and rather unspecific form, but this is a subject that deserves some attention and thought for the future. Indeed the idea is more or less futuristic and is more appropriate for the "electronic brains" of the future than for the computers now used, or under construction, or even planned. The basic idea is simple. Instead of having a single control unit sequencing the operations of the machine in series (except for certain subsidiary operations as certain input and output functions) as is now done, the idea is to decentralize control with several different control units capable of directing various simultaneous operations and interrelating them when appropriate.

We are still nowhere near the realization that Dr. Nash (then just J. Nash I suppose) had starting from the time when he proved the value of money to our society with "the bargaining problem". The letters to the NSA https://www.nsa.gov/public_info/_files/nash_letters/nash_letters1.pdf

The significance of this general conjecture, assuming its truth, is easy to see ... As ciphers become more sophisticated the game of cipher breaking by skilled teams, etc, should become a thing of the past.

He expresses a realization that didn't get (officially) implemented until Szabo posted the paper on the dawn of trustworthy computing.  Shortly after "the bargaining problem" and the letter to the nsa, comes the realization of Ideal Money, when fleeing to Europe and trying to exchange the USD for the Swiss currency (better quality). The Navy brought him back in chains.

All of these things together, create a machine and technology we still haven't come near realizing yet, either in the physical world or the mental, but the entire infrastructure has been mapped out for over 50 years, and a near complete design has been proposed for nearly 20. What we seemingly forget is this is the natural evolution of our civilization, AND these things have been in plain view the whole time....

"Satoshi" hides from the people, not the government. None of this is hidden, we just refuse to see.  Which of us has read all of this?  Are we sure this is insignificant and unrelated?  Are we grateful to Satoshi?

[traincarswreck]

So my personal view is that a practical global money might most favorably evolve through the development of a few regional currencies of truly good quality and then the "integration" or "coordination" of those global currency would become just a technical problem.

Here I am thinking of a politically neutral form of technological utility rather than of a money which for example, be used to exert pressures in a conflict situation comparable to "the cold war".

This variety of money would be intrinsically free of "inflationary decadence" similarly to how money was to be freer from that on a true "gold standard", but the proposed basis for that was not a proposal of a linkage to gold.

...it would be politically difficult to arrive at the implementation of such a system.

..the constitutional structure of the authority behind the euro is of the "paper money" kind ... But that is typical of all currencies used in the world nowadays

I dunno, you see there are some really really fun quotes here, and I don't know if we realize but eventually we are going to wake up one morning and collectively realize that all we even really needed to do to make bitcoin "work" is say this mans name, and this whole project will be the talk of the entire globe instantaneously:

Here are some more fun quotes:
https://www.youtube.com/watch?v=oH2bB7dYnAk

[iCEBREAKER]

Yes, a geometric increase is what I meant. Gavin's current proposal is for it to double every two years, meaning it increases 1.4X every year, which gives it a reasonably high chance of staying behind consumer bandwidth growth (and thus keeping the network highly decentralized).

So to answer your question, yes I would support something along the lines of what you're proposing. I think 5 MB would be much better than 2 MB though, because Bitcoin has a tendency to see sudden spurts in adoption, and so it'd be nice for it to have some room to grow quickly in the coming years.

Thanks for posting.

And I think one thing that often gets lost in the discussion on the hard limit, and that I can't stress enough:

There are other ways to stop bloat besides a hard limit in the protocol. A protocol limit is the most crude and inflexible way to counter spam/bloat. If a 40-50% increase per year ends up being faster than connection speed increases, it is very unlikely that other means will not be found to limit block sizes.

Yes, we have effective spam/bloat countermeasures.  That's why at present most blocks aren't full.

And Bitcoin certainly sees sudden spurts in adoption.  Thus my concern with the ultimate form of bloat: widespread actual usage.  

We need to understand how the system reacts to heavy actual usage.

Will anything break, or rather, what will degrade/break first?  How will the markets react?  What can be optimized and/or substituted given proper incentives such as the removal of free riders and their subsidized blockchain space?

It's nice we agree on a geometric increase, but wouldn't it be great to have actual data on which to better determine the optimum initial increase and eventual rate of increase?

Before changing the max_blocksize constant, we should know what happens to the BTC function at (and over) the 100% limit of the tx/block variable.

[CoinCidental]

Yes, a geometric increase is what I meant. Gavin's current proposal is for it to double every two years, meaning it increases 1.4X every year, which gives it a reasonably high chance of staying behind consumer bandwidth growth (and thus keeping the network highly decentralized).

So to answer your question, yes I would support something along the lines of what you're proposing. I think 5 MB would be much better than 2 MB though, because Bitcoin has a tendency to see sudden spurts in adoption, and so it'd be nice for it to have some room to grow quickly in the coming years.

Thanks for posting.

And I think one thing that often gets lost in the discussion on the hard limit, and that I can't stress enough:

There are other ways to stop bloat besides a hard limit in the protocol. A protocol limit is the most crude and inflexible way to counter spam/bloat. If a 40-50% increase per year ends up being faster than connection speed increases, it is very unlikely that other means will not be found to limit block sizes.

It's nice we agree on a geometric increase, but wouldn't it be great to have actual data on which to better determine the optimum initial increase and eventual rate of increase?

Before changing the max_blocksize constant, we should know what happens to the BTC function at (and over) the 100% limit of the tx/block variable.

we already know what will happen

 transactions wont confirm and there will be a backlog ,it would render the currency unusable if we had to wait hours or days to get included in a block

localbitcoins would die ,retailers would stop accepting btc for coffee or food etc because they cant force a person to wait all day until the transaction gets a confirmation

miners would include the highest paid transactions so it would mean higher fees for everyone to use btc efficiently   and the price would be all over the place  if you accepted $1000 in btc for a laptop it might be worth $900 by the time the transaction gets from the customers wallet to your wallet to the exchanges wallet etc

this would be huge to giant retailers

bottom line is we dont need to wait until it breaks and have a mad panic to get it fixed

do you drive  your car until the tyres explode  or replace them before that happens ?

[davout]

we already know what will happen

There is no "we", and you do not know.

[iCEBREAKER]

Before changing the max_blocksize constant, we should know what happens to the BTC function at (and over) the 100% limit of the tx/block variable.

we already know what will happen

 transactions wont confirm and there will be a backlog ,it would render the currency unusable if we had to wait hours or days to get included in a block

localbitcoins would die ,retailers would stop accepting btc for coffee or food etc because they cant force a person to wait all day until the transaction gets a confirmation

miners would include the highest paid transactions so it would mean higher fees for everyone to use btc efficiently   and the price would be all over the place  if you accepted $1000 in btc for a laptop it might be worth $900 by the time the transaction gets from the customers wallet to your wallet to the exchanges wallet etc

this would be huge to giant retailers

bottom line is we dont need to wait until it breaks and have a mad panic to get it fixed

do you drive  your car until the tyres explode  or replace them before that happens ?

I trust both my car's and tyres' manufacturers test every model to find out exactly how gracefully or suddenly their products' safety features degrade in the event of predictable and inevitable events like tyre explosions.  Stress-testing is Engineering 101.

Localbitcoins and retailers can easily increase their now-miniscule fees to ensure sufficient priority.

Miners would benefit from tx spaces being sold at the highest price the market will bear.

As fees increase, the more time and/or price sensitive among us will increasingly make use other of blockchains, as we should.

Bitcoin only needed to give out free samples long enough for word to spread.  It has.

Now is the best time to find out how the market for blockchain real estate responds to scarcity.

Indestructible, useful, eternal blockchain data storage is the most valuable service on earth; it must be optimized, not subsidized.

[traincarswreck]

This is "kinda" why I won't go away yet...because I showed through a new "variable" called "effective rake" (poker players were previously and largely unaware of), that ipoker (or other gaming in general) functions like that of an instable e-currency, or in other words poker sites print chips like the "keynesian" system Nash outlines in "ideal money".  The significance of this is multi-dimensional, and of many things is suggests it (poker and specifically "effective rake") can be stabilized much like our global economy by releasing in to the system a universal poker coin with the same printing rate as bitcoin.

Now I realize if you don't see bitcoin from the perspective I was introduced to it, Ideal Money, then this concept would sound completely foreign and possibly ludicrous.  But by the same measures bitcoin draws its value from the inefficiencies of our "Keynesian" banking system, our universal poker coin will draw out the "effective rake" of the ipoker industry.

Poker players are the most densely connected socially network group of payment processor savy (except crypto so far, but largely because sites have incentive to downplay the utility) players that whoreship the strategy math laid out by John Nash. The average online tournament player uses this mans name multiple times a day discussing strategy with peers.  The coin we have released will hit its tipping point in adoption and create a fury of cooperative game theoretical strategy discussions in relation to Ideal Poker and its relation to Ideal Money and its relation to bitcoin...

And we should likely then see a 50+billion dollar industry flip to crypto nearly overnight (if we believe that both the price of bitcoin and the price players universal poker coin would incentivize such change)...

So I see a different plan and a different set up.  But here some interesting observations:

1) I'm clearly the most well read on these related subjects (easy because nobody is reading/talking about ideal money).
2) The coin I present, is more decentralized and more owned by the people and the citizens and not a private third party than any other coin you will find proposed.
3) Ideal Poker and the plan I present was described over a year ago before I knew what bitcoin was and certainly before bitcoin 2.0 (ie counterparty).

Cliffs: Poker will move first, as it should, because "densely connected population game theorists"

[davout]

So much fucking word salad....

[CoinCidental]

Before changing the max_blocksize constant, we should know what happens to the BTC function at (and over) the 100% limit of the tx/block variable.

we already know what will happen

 transactions wont confirm and there will be a backlog ,it would render the currency unusable if we had to wait hours or days to get included in a block

localbitcoins would die ,retailers would stop accepting btc for coffee or food etc because they cant force a person to wait all day until the transaction gets a confirmation

miners would include the highest paid transactions so it would mean higher fees for everyone to use btc efficiently   and the price would be all over the place  if you accepted $1000 in btc for a laptop it might be worth $900 by the time the transaction gets from the customers wallet to your wallet to the exchanges wallet etc

this would be huge to giant retailers

bottom line is we dont need to wait until it breaks and have a mad panic to get it fixed

do you drive  your car until the tyres explode  or replace them before that happens ?

I trust both my car's and tyres' manufacturers test every model to find out exactly how gracefully or suddenly their products' safety features degrade in the event of predictable and inevitable events like tyre explosions.  Stress-testing is Engineering 101.

stress testing is exactly that and done under controlled conditions to find out the breaking point
you dont stress test the only working version that everyone is currently using

if you want to make a stress test ,do it with a valueless altcoin or make a new coin just for that purpose

like how scientists test new drugs on animals BEFORE giving  them to the public

[Buffer Overflow]

And it's only 3 transactions per second, not 7.

Ok lol.

I would err on the side of caution and say 2tps myself.

A good estimate for the network throughput when limited to 1MB blocks would be 2 to 4 tps depending on how optimistic (naive) you want to be.

[Rampion]

Why don't we just double the block size limit when we approach it? We could do it every time the limit is approached.

That seems more sensible than just making it x20 in one go.