I always liked this one.
Happy Birthday Woodcoin!
Topic: Happy Birthday Woodcoin - Reflections on two years chopping LOG (Read 290 times)
October 20, 2016, 10:02:03 AM
Two Years of Chopping LOG
http://frass.woodcoin.org/woodcoin-two-years-of-chopping-log/
Happy birthday Woodcoin!
I'm going to use this excuse to take a look at the current state of the network and review the motivation behind using woodcoin.
Part 1 PRICE
Fig. 1 Market Capitalization of Woodcoin over first 2 years.
Woodcoin is a small coin listed with a capitalization of ~200BTC. This puts it at the 109th largest "mineable and non-premined" coin listed on coinmarketcap. As you probably know by looking at this metric before, it is easily skewed by small distributions of stackers, small volume, and coins burned or lost. I expect many of the coins in this top 100 list to disappear just like paycoin did. The half a million LOG sent to the address WeHonorTheForestsAndTheTrees4pPXTQ are not included here so really the spendable coins are at least 6% smaller than reported.
Woodcoin is fully public and decentralized with no ICO, no registered foundation or corporation, nor any solicitation or acceptance of fiat currencies during its creation. If you squint a bit you can see this in the charts. One immediately notices in Figure 1 that the usual fractal pattern exhibited in most coins (large fast pumps followed by slow decreases) are not visible here. Where are the pump and dumps? Where is the ICO spike? Woodcoin shows a stable (but slow) growth, which is no surprise to those who understand the supply curve (see details below). Of course because woodcoin is open to all, this is no guarantee of future behavior.
While woodcoin is currently listed on at least four currently operating exchanges, most all of the volume is on C-CEX. They have proved to be reliable and trustworthy custodians as of today, and have also shown consistent growth over the last two years. You know the deal with exchanges by now.
It's also worth pointing out that C-CEX reported LOG trading for as high as 8000 sat (during one emptying of supply about six months ago) and a couple of faucets currently sell them in small quantities for 0 sat. That should give you an idea of the ranges of prices.
The total supply (including permanently tied LOG) is now just over 8 million LOG. This is out of a maximum of 27.6 million LOG, which as you probably know, will take a long time to chop
Part 2 HASH RATE
Fig. 2 Hash rate of LOG choppers over the first two years of woodcoin.
The constant battle which is a proof-of-work network has been very exciting to watch here. Recently (as in yesterday) the difficulty has touched all time highs. There are quite a lot of skein hash functions being calculated, to say the least. The network is currently in the GPU phase of growth, which is to say that the vast majority of the hashpower appears to be in GPU rigs.
There have also been hash withdrawl "attacks" which began almost a year ago, in November of 2015. Certain GPU farmers dedicated their relatively large rigs to chopping LOG for short time periods, then left the network when the difficulty increased, presumably to mine other coins. This led to periods of very long block times. LOG users waited patiently and the network recovered as it should. While some people reported being annoyed by long confirmation times, all transactions went through. In addition to complaints there was much discussion on what could be changed (with a fork) to alter these dynamics. A recent talk by Mark Friedenbach at Scaling Bitcoin in Milan touched on these issues. Difficulty adjust algorithms such as the one employed by woodcoin look dreadfully simple to a control systems engineer (why is there no PID controller?), but their simplicity and resilience make them work - even if at times their overshoots and undershoots are aggravating.
This summer further hash warfare activity was detected on the network, in the form of block withholding "attacks". These enabled certain woodcutters to re-chop blocks of LOG (which had been chopped by hash-withrdawl players as discussed above) after the original hashpower was withdrawn. This caused a certain amount of confusion but good will was displayed by all involved as no transactions were reported double-spent, despite indications that some woodcutters had the power to perform a dreaded 51% attack. The woodcutters involved even went so far as to generously refund any smaller choppers who had lost their rewards as orphans. This led one woodcoin user to ask me "Why is it raining LOGs?" There were even one or two chain reorganizations that numbered as high as thousands of blocks. While hardly a selling point on the security of woodcoin transactions at that moment, this was very exciting to see that the consensus network can recover from such shenanigans. Small and decentralized proof of work chain dynamics can be very exciting.
Also related was a brief "Quantitative Easing" period sponsored by the secretive dwarven Kawg trust, in which large transaction fees were paid to woodcutters. There are rumors that another period of quantitative easing or QE2 will emerge on the network.
Overall the network has behaved well, which even after 8 years of watching bitcoin blocks appear still seems somewhat of a miracle. Lets not forget this was proven by some of the very best cryptographers to be impossible
Part 3 MOTIVATION
The real motivation is for our betters, the trees, to survive and thrive, and for us to secure our life support and to gain a second spaceship to ensure survival of Gaia. We fight Sauron's colonial empire within as well as the crawling radiative chaos Nyarlathotep without. That's a long story we won't get into here. To get started with that, one thing we need is to remove the gaping wound on humanity and middle earth which reliance on private coin issuance has opened, and to do that we need a strong ecosystem of public coins. Woodcoin aims to be one of these coins.
When I presented the logarithmic supply curve, and implemented it with the woodcoin genesis block two years ago, I didn't go out of my way to explain it. Some people figured it out anyway.
The problem we are solving here has been largely overlooked, but it hasn't gone away. Namely: What is going to incentivise miners on the bitcoin (or other geometric supply curve coin) network when the block subsidy reward drops to near zero? Satoshi's answer, and perhaps the only answer that makes any sense, is transaction fees. However, we have a large segment of bitcoin users who fear that large transaction fees will push users to other coins. Some even suggest raising the maximum block size to avoid these large fees. In addition to this big question mark of how the network will work when block subsidies drop to near zero, as they do rapidly in a geometric release curve, this rapid release makes the supply look unattractive to new users for another reason: it appears as a premine and as an unfair distribution. Currently more than three quarters of all bitcoin to ever be mined are already in the pockets of early adopters. Many critiques of BTC harp back to this point. [Half the LOG will be released by 2305. Three quarters won't be hit for thousands of years after that, I'll let you do the calculation.]
One potential way around these problems which has the advantage of simplicity is to have a constant inflation. This is the solution of dogecoin and of ethereum: an eventual minimum non-shrinking block subsidy. This plan has its attraction, but it also might scare new users away because it means that the money supply is not capped. As time crawls on, value in a single token thus inevitably goes down, and long term store-of-value seekers flee. Another uncapped strategy is that of many proof-of-stake coins, namely an constant percentage reward, leading to a still faster exponential supply growth.
Fig. 3 Three existing supply curves, taken from Margrit Kennedy "Interest and Inflation Free Money". Curve A is a geometric release, like bitcoin or litecoin. Curve B is a constant subsidy, like doge or ethereum. Curve C is a constant percentage inflated release like most proof of stake coins.
The logarithmic supply takes a middle road between the geometric release curve and the constant release curve (A and B in figure 3). Unlike a constant release curve, the supply is capped - there is a limit which will never be passed (technically due to the discrete nature of the supply, in continuous fields logarithms are unbounded). However it approaches this limit incredibly slowly which ensures there will be some subsidy to support securing the chain. Further, there is always an advantage to securing the network today, as compared to securing it tomorrow. Every block is worth less than its prior.
Part 4 IMPLEMENTATION
Once you understand the purpose of the logarithmic release function, the rest of the woodcoin design mostly follows. The satoshi codebase was the choice of instantiation, due to the relatively well tested nature and the existing tools to work with it. Several clients have emerged, as well as tools for address manipulation and so forth. Simplicity is an overall goal here, as we want people to be able to use this just as easily as any other public coin. Design consideration was conservative, as recommended here.
Once the supply curve was decided, the big question was what to use for a PoW hash. It's not a good idea to use a hash function that other big coins are already using (for background see why Doge is now mergemined on litecoin or why NMC is now mergemined on Bitcoin). At the time, Quarkcoin had just come out and people were excited about chaining together hash functions to make it more difficult to do custom hardware. As I saw it, custom hardware was an eventuality for any coin if it lived long enough to become popular, and chaining hash functions only made the thing more complex and I saw it as reducing the value. At the time SHA3 was just being chosen, and one of the potential candidates was the Skein function. Bruce Schneier and three fish told me it worked, the NSA rejected it, some research suggested it was fast and would be relatively easy to optimize in custom hardware, and so I went with that. I wasn't about to roll my own hash function yet, sorry [It's worth pointing out that woodcoin is the current leader in pure Skein hashrate].
The only other major thing different about woodcoin is the elliptic curve. I've written about this before so I'll spare you the discussion again. I went with something different than bitcoin because it was easy, well tested, and recommended by many of the experts. Conservatism again, boring but that's the way it is.
The rest, as they say, is history.
Part 5 ACKNOWLEDGEMENTS
Many thanks to all the great folks I have met through this project! Keep up the good work in all endeavors, public coin and otherwise.
http://frass.woodcoin.org/woodcoin-two-years-of-chopping-log/
Happy birthday Woodcoin!
I'm going to use this excuse to take a look at the current state of the network and review the motivation behind using woodcoin.
Part 1 PRICE
Fig. 1 Market Capitalization of Woodcoin over first 2 years.
Woodcoin is a small coin listed with a capitalization of ~200BTC. This puts it at the 109th largest "mineable and non-premined" coin listed on coinmarketcap. As you probably know by looking at this metric before, it is easily skewed by small distributions of stackers, small volume, and coins burned or lost. I expect many of the coins in this top 100 list to disappear just like paycoin did. The half a million LOG sent to the address WeHonorTheForestsAndTheTrees4pPXTQ are not included here so really the spendable coins are at least 6% smaller than reported.
Woodcoin is fully public and decentralized with no ICO, no registered foundation or corporation, nor any solicitation or acceptance of fiat currencies during its creation. If you squint a bit you can see this in the charts. One immediately notices in Figure 1 that the usual fractal pattern exhibited in most coins (large fast pumps followed by slow decreases) are not visible here. Where are the pump and dumps? Where is the ICO spike? Woodcoin shows a stable (but slow) growth, which is no surprise to those who understand the supply curve (see details below). Of course because woodcoin is open to all, this is no guarantee of future behavior.
While woodcoin is currently listed on at least four currently operating exchanges, most all of the volume is on C-CEX. They have proved to be reliable and trustworthy custodians as of today, and have also shown consistent growth over the last two years. You know the deal with exchanges by now.
It's also worth pointing out that C-CEX reported LOG trading for as high as 8000 sat (during one emptying of supply about six months ago) and a couple of faucets currently sell them in small quantities for 0 sat. That should give you an idea of the ranges of prices.
The total supply (including permanently tied LOG) is now just over 8 million LOG. This is out of a maximum of 27.6 million LOG, which as you probably know, will take a long time to chop
Part 2 HASH RATE
Fig. 2 Hash rate of LOG choppers over the first two years of woodcoin.
The constant battle which is a proof-of-work network has been very exciting to watch here. Recently (as in yesterday) the difficulty has touched all time highs. There are quite a lot of skein hash functions being calculated, to say the least. The network is currently in the GPU phase of growth, which is to say that the vast majority of the hashpower appears to be in GPU rigs.
There have also been hash withdrawl "attacks" which began almost a year ago, in November of 2015. Certain GPU farmers dedicated their relatively large rigs to chopping LOG for short time periods, then left the network when the difficulty increased, presumably to mine other coins. This led to periods of very long block times. LOG users waited patiently and the network recovered as it should. While some people reported being annoyed by long confirmation times, all transactions went through. In addition to complaints there was much discussion on what could be changed (with a fork) to alter these dynamics. A recent talk by Mark Friedenbach at Scaling Bitcoin in Milan touched on these issues. Difficulty adjust algorithms such as the one employed by woodcoin look dreadfully simple to a control systems engineer (why is there no PID controller?), but their simplicity and resilience make them work - even if at times their overshoots and undershoots are aggravating.
This summer further hash warfare activity was detected on the network, in the form of block withholding "attacks". These enabled certain woodcutters to re-chop blocks of LOG (which had been chopped by hash-withrdawl players as discussed above) after the original hashpower was withdrawn. This caused a certain amount of confusion but good will was displayed by all involved as no transactions were reported double-spent, despite indications that some woodcutters had the power to perform a dreaded 51% attack. The woodcutters involved even went so far as to generously refund any smaller choppers who had lost their rewards as orphans. This led one woodcoin user to ask me "Why is it raining LOGs?" There were even one or two chain reorganizations that numbered as high as thousands of blocks. While hardly a selling point on the security of woodcoin transactions at that moment, this was very exciting to see that the consensus network can recover from such shenanigans. Small and decentralized proof of work chain dynamics can be very exciting.
Also related was a brief "Quantitative Easing" period sponsored by the secretive dwarven Kawg trust, in which large transaction fees were paid to woodcutters. There are rumors that another period of quantitative easing or QE2 will emerge on the network.
Overall the network has behaved well, which even after 8 years of watching bitcoin blocks appear still seems somewhat of a miracle. Lets not forget this was proven by some of the very best cryptographers to be impossible
Part 3 MOTIVATION
The real motivation is for our betters, the trees, to survive and thrive, and for us to secure our life support and to gain a second spaceship to ensure survival of Gaia. We fight Sauron's colonial empire within as well as the crawling radiative chaos Nyarlathotep without. That's a long story we won't get into here. To get started with that, one thing we need is to remove the gaping wound on humanity and middle earth which reliance on private coin issuance has opened, and to do that we need a strong ecosystem of public coins. Woodcoin aims to be one of these coins.
When I presented the logarithmic supply curve, and implemented it with the woodcoin genesis block two years ago, I didn't go out of my way to explain it. Some people figured it out anyway.
The problem we are solving here has been largely overlooked, but it hasn't gone away. Namely: What is going to incentivise miners on the bitcoin (or other geometric supply curve coin) network when the block subsidy reward drops to near zero? Satoshi's answer, and perhaps the only answer that makes any sense, is transaction fees. However, we have a large segment of bitcoin users who fear that large transaction fees will push users to other coins. Some even suggest raising the maximum block size to avoid these large fees. In addition to this big question mark of how the network will work when block subsidies drop to near zero, as they do rapidly in a geometric release curve, this rapid release makes the supply look unattractive to new users for another reason: it appears as a premine and as an unfair distribution. Currently more than three quarters of all bitcoin to ever be mined are already in the pockets of early adopters. Many critiques of BTC harp back to this point. [Half the LOG will be released by 2305. Three quarters won't be hit for thousands of years after that, I'll let you do the calculation.]
One potential way around these problems which has the advantage of simplicity is to have a constant inflation. This is the solution of dogecoin and of ethereum: an eventual minimum non-shrinking block subsidy. This plan has its attraction, but it also might scare new users away because it means that the money supply is not capped. As time crawls on, value in a single token thus inevitably goes down, and long term store-of-value seekers flee. Another uncapped strategy is that of many proof-of-stake coins, namely an constant percentage reward, leading to a still faster exponential supply growth.
Fig. 3 Three existing supply curves, taken from Margrit Kennedy "Interest and Inflation Free Money". Curve A is a geometric release, like bitcoin or litecoin. Curve B is a constant subsidy, like doge or ethereum. Curve C is a constant percentage inflated release like most proof of stake coins.
The logarithmic supply takes a middle road between the geometric release curve and the constant release curve (A and B in figure 3). Unlike a constant release curve, the supply is capped - there is a limit which will never be passed (technically due to the discrete nature of the supply, in continuous fields logarithms are unbounded). However it approaches this limit incredibly slowly which ensures there will be some subsidy to support securing the chain. Further, there is always an advantage to securing the network today, as compared to securing it tomorrow. Every block is worth less than its prior.
Part 4 IMPLEMENTATION
Once you understand the purpose of the logarithmic release function, the rest of the woodcoin design mostly follows. The satoshi codebase was the choice of instantiation, due to the relatively well tested nature and the existing tools to work with it. Several clients have emerged, as well as tools for address manipulation and so forth. Simplicity is an overall goal here, as we want people to be able to use this just as easily as any other public coin. Design consideration was conservative, as recommended here.
Once the supply curve was decided, the big question was what to use for a PoW hash. It's not a good idea to use a hash function that other big coins are already using (for background see why Doge is now mergemined on litecoin or why NMC is now mergemined on Bitcoin). At the time, Quarkcoin had just come out and people were excited about chaining together hash functions to make it more difficult to do custom hardware. As I saw it, custom hardware was an eventuality for any coin if it lived long enough to become popular, and chaining hash functions only made the thing more complex and I saw it as reducing the value. At the time SHA3 was just being chosen, and one of the potential candidates was the Skein function. Bruce Schneier and three fish told me it worked, the NSA rejected it, some research suggested it was fast and would be relatively easy to optimize in custom hardware, and so I went with that. I wasn't about to roll my own hash function yet, sorry
[It's worth pointing out that woodcoin is the current leader in pure Skein hashrate].The only other major thing different about woodcoin is the elliptic curve. I've written about this before so I'll spare you the discussion again. I went with something different than bitcoin because it was easy, well tested, and recommended by many of the experts. Conservatism again, boring but that's the way it is.
The rest, as they say, is history.
Part 5 ACKNOWLEDGEMENTS
Many thanks to all the great folks I have met through this project! Keep up the good work in all endeavors, public coin and otherwise.
Jump to: