The single best resource seems to be https://en.bitcoin.it/wiki/Contracts but it only covers a fraction of the possibilities; for example, I corresponded with Hearn about timestamping for time-lock crypto but that hasn’t been added.
The single best resource seems to be https://en.bitcoin.it/wiki/Contracts but it only covers a fraction of the possibilities; for example, I corresponded with Hearn about timestamping for time-lock crypto but that hasn’t been added.
Excellent! I had no idea the bitchain included all those extra fields. So much potential!
Yeah; people don’t realize how Bitcoin can do so much. It’s not (just) sending some irreversible transactions around, as useful as it is to replace Paypal/WU. It’s like every financial cryptographic construct ever got mashed up in an academic paper, except it’s not a paper, it’s a real system in use worldwide and the constructs can be actually used by real people for real things. Of course none of them are easy to use, many aren’t even enabled—but I think the situation is like the early Linux kernel: it’s a rare geek who understands why you would want to hack on a kernel, a rarer techie who can compile kernels and mess with kernel settings, and an even rarer techie who can write a kernel module to do something, but the potential and power are there.
Of course none of them are easy to use, many aren’t even enabled—but I think the situation is like the early Linux kernel: it’s a rare geek who understands why you would want to hack on a kernel, a rarer techie who can compile kernels and mess with kernel settings, and an even rarer techie who can write a kernel module to do something, but the potential and power are there.
You have just created “a rare geek who understands why you would want to hack on a kernel”. These features are screaming out to be used and may facilitate some of the projects that I have in mind (including some plans for small-scale crowdfunded research).
Do you have any other resources on bitcoin to recommend? In particular I a resource that lists definitively which features are enabled and ready to use in current clients? (I hate accidentally building or planning things that rely on features not yet in the mainstream software. It feels like crippling myself when I try to backport.)
Do you have any other resources on bitcoin to recommend? In particular I a resource that lists definitively which features are enabled and ready to use in current clients?
No idea. I have a vague understanding that colored coins are a work in progress and may be added at some point in 2013, but I’ve never seen anyone lay out a list of the scripting language elements with the enabled and disabled ones specified. Since this is such an obscure area of Bitcoin, you probably will have to read the source to find out.
Since this is such an obscure area of Bitcoin, you probably will have to read the source to find out.
Time for me to do an expected value calculation. Is it worth becoming an expert on bitcoin? Not at all implausible. Apart from the intrinsic value of personal interest there seem to be things that the world needs that bitcoin could help with (an actual prediction market that doesn’t suck for example). Some things that the world needs it is possible to create ways to be paid for doing. It aligns well with my aptitudes.
Creating a bitcoin prediction market is a great idea.
One thing I’ve been wondering recently is why more people don’t create Bitcoin alternatives. Many of the alternatives that already exist have ridiculously high market capitalizations, with what appears to be minimal marketing and almost zero rationale for their existence. If you could create an alternative that had “baked in” solutions to some of the problems people complain about re: long-term adoption of bitcoin (limited divisibility, difficulty of transferring in/out of USD, volatility (somehow fix this? e.g. if the protocol was somehow robust to high-frequency trading, perhaps that would deal with volatility as HFT people traded using volatility pumping strategies), performance problems as the network grows large, general awkwardness of waiting for confirmations and stuff), or had some kind of killer app that bitcoin didn’t have, and then put an actual marketing push behind all this (and bribe the existing exchanges to accept it or find a way to piggyback on the existing bitcoin infrastructure), it seems like it could be a great way to make money.
Creating a bitcoin prediction market is a great idea.
I’m considering seriously. In the last several days I’ve analyzed the technology already in widespread, looked at the possibilities and explored the limits. The potential is huge. Cryptography is like wizardry that works in the real (abstract) world!
One thing I’ve been wondering recently is why more people don’t create Bitcoin alternatives.
The main not-ridiculously-pointless alternative on the horizon is ripple, which has a different focus to bitcoin. If they have marketing awesomeness behind them they could get people to play their game. (Low probability, high reward for them.)
An awesome cryptographic feature of the bitchain mining system is that new bitchains can be mined using the same computational resources that bitcoin miners use for negligible extra effort. ie. Instead of creating lottery tickets and entering them in one lottery they can create lottery tickets and enter the same ticket in 20 different lotteries. This drastically lowers (but doesn’t remove) the difficulty in getting alternate bitchains to be secure.
If you could create an alternative that had “baked in” solutions to some of the problems people complain about re: long-term adoption of bitcoin (limited divisibility
Is this a significant problem? 1 satoshi (the minimum division) is 0.000 000 01 BTC. That’s… not very much. And from one I understand changing the system such that it could allow lower divisions would be plausible if it became necessary. (ie. It is the kind of change that could easily be implemented and adopted by a supermajority of computing power of the miners and thereby become “distributed official”.)
, difficulty of transferring in/out of USD
The ‘ripple’ currency system has dreams of solving this via webs of (limited) trust and social networks. If this were solved in the ripple network it would also incidentally solve it in any currency that is significantly used.
, volatility (somehow fix this? e.g. if the protocol was somehow robust to high-frequency trading, perhaps that would deal with volatility as HFT people traded using volatility pumping strategies),
This volitilty can be solved within bitcoin itself via the adoption of “coloured coins”. Essentially, this means credible institutions can issue “MyBankUSDcoins” for example, with coins with a certain history being declared redeemable at a specified rate via whatever banking mechanism that kind of company does. Similarly pecunix could offer “Pecunix coins” which it backs with gold (its normal business model). Then transactions can be made in gold, or USD, or AUD via the same technology (and distributed security structure) that bitcoin uses.
The above is possible right now. It requires no change to the bitcoin (mining server side) software. It is just a matter of people with sufficient social influence potential making it happen and customers (including merchants) being convinced that they want it.
performance problems as the network grows large, general awkwardness of waiting for confirmations and stuff)
Yes, systems without the awkwardness would have an advantage.
Has anyone tried to solve the marketing issue by giving their coin a referral system similar to what Paypal had in the early days? I’m not 100% sure about this, but I think maybe the way it worked was if you referred your friend to Paypal, you would get $x and your friend would get $x. Supposedly this was a big reason why Paypal beat out competing services in its early days, ’cause people had this monetary incentive to spread it. (I have an acquaintance who worked for a startup that competed with Paypal and lost; I could probably get in contact with him if anyone has any questions.)
This volitilty can be solved within bitcoin itself via the adoption of “coloured coins”. Essentially, this means credible institutions can issue “MyBankUSDcoins” for example, with coins with a certain history being declared redeemable at a specified rate via whatever banking mechanism that kind of company does. Similarly pecunix could offer “Pecunix coins” which it backs with gold (its normal business model). Then transactions can be made in gold, or USD, or AUD via the same technology (and distributed security structure) that bitcoin uses.
Hm. Sounds like colored coins have all the advantages of Bitcoin, plus lack of volatility. So if they existed then would we ditch regular Bitcoins for them? How much software would need to be changed in order to implement colored coins?
Ultimately, I suspect that Bitcoin is a product that no one really wants that bad, and ultimately the interest in them is mostly from speculation… if there’s any interest in using them as an actual currency, it’s liable to be faked interest from people who have large bitcoin balances and would profit if that happened.
Ultimately, I suspect that Bitcoin is a product that no one really wants that bad
Well there was a lot of interest in Bitcoins in southern Europe following the Cyprus, bank account confiscations. Basically, if people are treating putting money into Bitcoins as an alternative to putting it in banks.
Hm. Sounds like colored coins have all the advantages of Bitcoin, plus lack of volatility.
Alas no. They have most of the advantages of bitcoin since they are still bitcoin in full. However, whatever additional value the ‘color’ grants to the coin over the base worth os a BTC is based on trust in a specific agent. They are nothing more than a promise from a supplier to redeem the coins into another form of value at some time in the future. It is a trade off. Convenience and stability purchased by requiring trust.
So if they existed then would we ditch regular Bitcoins for them?
The regular coins are still what are required to produce coloured coins (the name is really good analogy). Coloured coins being used more would seem to lower the expected price of uncoloured bitcoin, but they couldn’t make uncouloured coins worthless since they themselves entail a demand for them.
How much software would need to be changed in order to implement colored coins?
It would mean taking the existing ‘beta’ implementations, making them stable and having people start to use them. Merchants accepting them automatically (and being able to tell they are worth more than uncouloured coins) would also require making improvements to whatever software it is that they use now.
An awesome cryptographic feature of the bitchain mining system is that new bitchains can be mined using the same computational resources that bitcoin miners use for negligible extra effort. ie. Instead of creating lottery tickets and entering them in one lottery they can create lottery tickets and enter the same ticket in 20 different lotteries. This drastically lowers (but doesn’t remove) the difficulty in getting alternate bitchains to be secure.
I thought merged mining only worked for Bitcoin+Namecoin? I was looking into mining Bitcoin and Litecoin simultaneously, but it looked like it simply wouldn’t work with cgminer and in general.
I thought merged mining only worked for Bitcoin+Namecoin? I was looking into mining Bitcoin and Litecoin simultaneously, but it looked like it simply wouldn’t work with cgminer and in general.
Merged mining requires a small change to the bitchain of all but one of the currencies. So you can merge mine bitcoin + 20 merge mining enabled currencies OR you can merge mine litecoin + 20 merge mining enabled currencies.
Currencies that can currently be merge mined with BTC (and that I am aware of):
NameCoin—Store aliases and soforth conveniently. Is complementary to bitcoin for some use-cases.
DevCoin --- (Roughly speaking) a charity for funding opensource developers.
TonalCoin—Which somehow is also embedded in the bitcoin blockchain and allows bitcoins to be traded using a different number system.
IxCoin—Utterly pointless attempt to cash in on crypto speculation.
No idea. I have a vague understanding that colored coins are a work in progress and may be added at some point in 2013
I have been looking into the protocol for how this works. It doesn’t require anything additional to be added (to the bitcoin daemon). To be practical it does require that clients are made that process them conveniently, this is what is the work in progress and does seem to be at the ‘usable in a stable form some time this year’ point.
The above distinction matters to me because it influence whether it is something I can just use if I find customers who want it and I am willing to do some coding to facilitate it versus features that require creation of a Bitcoin Improvement Proposal then the politically influence of a miners to start adopting it. For example zerocoin (the solution to anonymity and privacy) would require huge political advocacy to get into play in the existing bitcoin currency.
but I’ve never seen anyone lay out a list of the scripting language elements with the enabled and disabled ones specified.
Indeed, my research has failed to turn up such a document too. The information that is out there is also significantly out of date unless pieced together rather carefully. A summary of the availability of the features that are of interest. I am reasonably confident in these findings but not certain. (90% confidence.)
Timelocked transactions: The protocol (and actually in use bitchain) supports it but miners will nearly always reject transactions with future dates. This seems stupid to me but there doesn’t seem to be any strong pressure to start accepting them. It is still possible to create such transactions and store them until the time is up then send them in. This solves some timelock usecases but the possibility of double spending prevents it solving many other use cases.
Multisignature transactions. ie. Escrow. Transactions that require n of m signatures in order to spend the output. These work. Technically. You can create them manually via the command line, send them to the blockchain and then use the command line and use of or access to someone who will cooperate with n of m signatures to create transactions that spend them. But current clients don’t do this for you (except in the crippled case whereby if you have m of m keys in your wallet it will deign to let you spend them.) But this feature is there to be used if you want it.
Zero Knowledge Proofs eg. Zerocoins. Reliable privacy and anonymity right there within bitcoin. The research and cryptographic technology is available. As are proof of concept reference implementations. But this system (or an equivalent or superior one) isn’t accepted and likely will not be for at least several years. An alternate currency could do it today.
Since this is such an obscure area of Bitcoin, you probably will have to read the source to find out.
Done. Now I’ll have to read the source even more to work out how to make clients that use the (available) features to serve my purposes.
Well, good luck. I don’t know a lick of C++ or real crypto coding, so as interesting as I find it all to read and watch and write essays about, don’t expect any help from me. I hope you succeed in getting some of the features in more general use.
But this system (or an equivalent or superior one) isn’t accepted and likely will not be for at least several years. An alternate currency could do it today.
But also risks there being some minor problem with the unpublished Zerocoin proposal which de-anonymizes everyone who ever uses the functionality. Since laundries/mixes currently seem to be working pretty well, it’s better to be conservative and get it right, than push for it to be implemented right now and in 2 years, watch every transaction be de-anonymized.
Yeah; people don’t realize how Bitcoin can do so much. It’s not (just) sending some irreversible transactions around, as useful as it is to replace Paypal/WU.
On a related note do you have any thoughts on ripple? It seems to offer huge potential convenience if ‘distributed payment processing’, which could work well to enable bitcoin to be easily spent and traded. From what I can tell their business model is based on making an essentially free, incredibly convenient service and if it takes off it will produce value in their own currency of which they will (allegedly) be distributing 80% for free. I don’t precisely understand how it magically operates as a distributed currency exchange and payment processor but it seems plausible (if enough people want it).
Reading about it right now, I would analogize it to an automated barter system, the hawala money transmission system, pre-modern banking’s system of drafts and correspondents like the Medici Bank, and Szabo’s bitgold minting.
If I’m understanding the wiki documentation right, you can basically think of Ripple as a Bitcoin in which everyone can issue as many bitcoins as they want but each user produces a different kind of bitcoin. The only reason anyone is willing to hold a wedrifid-coin is because they know you in real life and expect you to pay up, or the reason someone is willing to hold gwern-coins is because I have in the past upheld my agreements to, say, order something off Amazon.com for them. I might decide to accept wedrifid-coins and now the wedrifid-coin holders have the option to either shake you down in real life or buy something off Amazon via me; and if someone else decided to accept gwern-coins, a holder of wedrifid-coins could convert to gwern-coins and then convert to someoneelse-coins.… Automated barter of personal IOUs, almost—like those circles you hear of involving scores of people in which A gives something to B, B gives something to C, and so on to Z who gives A what A really wanted (most famously used in kidney transplants; I need a transplant, but my brother is not compatible with me, so he gives his kidney to B, and B’s son gives a kidney to C, and C’s father gives me a compatible kidney).
Since you’re deciding whose coins to accept directly or transitively, Ripple seems to pass the security buck to your decision there. This obviously makes Ripple’s job much much easier. Since you’ll be deciding basically on social or extra-network factors, Ripple also doesn’t bother with any worries about anonymity. (Their list of features re mining, if I’m understanding the architecture right, less than impressive in this light.)
There’s also a centralized aspect which is nasty and probably part of how they plan to monetize it.
So yeah, interesting. I don’t feel compelled to trade in my one bitcoin for any Ripple stuff, though.
I don’t precisely understand how it magically operates as a distributed currency exchange and payment processor but it seems plausible (if enough people want it).
I think it operates pretty much the same way a PGP web of trust operates. You specify a few public keys who you trust, and you can bootstrap from there.
If I’m understanding the wiki documentation right, you can basically think of Ripple as a Bitcoin in which everyone can issue as many bitcoins as they want but each user produces a different kind of bitcoin.
So (I assume) the way it can operate as a distributed currency exchange is if there is someone with Yen out there and someone with USD out there and either Yen-guy or USD-guy trust each other directly or their is a web of trust between them. So RandomJoe can sell Yen-guy coins for stuff and buy USD-coins then USD guy will give actual cash to him.
Yes, I think that’s how it could go down. Or Yen-guy and USD-guy could exchange currency directly, or they could do it hawala-style indirectly (for example: Yen-guy could hand out yen to anyone wandering into his Tokyo shop with USD-guy coins, and then USD-guy, to settle his bills with Yen-guy, can fly in and visit Yen-guy with a case of wine which Yen-guy can drink or sell).
I wonder if a form of coin-locking could be used to create simulated versions of other currencies. Depending on the demand and level of trust in the given currency, a given bitcoin sum could be locked in such a way that it gets automatically distributed to all holders at some future point (if they don’t choose to hold onto the simulated currency instead) in proportion to how much they have. The choice to hold or drop the simulated currency would be a way of measuring its value against the bitcoin.
I wonder if a form of coin-locking could be used to create simulated versions of other currencies. Depending on the demand and level of trust in the given currency, a given bitcoin sum could be locked in such a way that it gets automatically distributed to all holders at some future point (if they don’t choose to hold onto the simulated currency instead) in proportion to how much they have. The choice to hold or drop the simulated currency would be a way of measuring its value against the bitcoin.
I can think of various ways to facilitate this using the techniques I am considering (or considered and discarded) for the creation of a distributed prediction market. The details of the precise implementation would depend on precisely what your goal was. I can’t fully determine what you are aiming at from what you have said so far? More detail? A use case?
What I have in mind is something a little bit like the zerocoin proposal, in the respect that it’s an alternate coin which can be translated to Bitcoin and back as part of how the system works.
But instead of being fixed with respect to the bitcoin it would increase or decrease in relative value in response to market signals. There would be an exchange rate determined by dividing the number of spendable units of alternate currency against the number of bitcoins locked behind it.
So what you need is a series of rules for controlling that number that tend towards economic stability. Which is itself kind of tricky. I had thought locking the coins by a time-lock could do it, but I haven’t thought of a way to base an exchange rate on that. Empirical work with simulator or a smart math person with a whiteboard might turn up something.
The following uses incentives to accomplish something similar-in-principle:
Alice creates 100 dollarcoins at a cost of 1 bitcoin. She transfers 1 to Bob, he can turn it into 0.01 bitcoin.
The dollarcoin has a specification that includes things like interest payments. If Bob waits 30 days after receiving a dollarcoin he gets 1% of its nominal bitcoin value in interest. He gets another 1% payment 30 days later, and so forth. But if he spends it right away or converts it to bitcoin there are no payments.
Claire decides to use the dollarcoin specification to mint some dollarcoins. She has to pay a 1% minting fee, so 1.01 bitcoins to make 100 dollarcoins. The backing rate becomes 200 dollarcoins to 2.01 bitcoins.
Thus the more dollarcoins get minted the more valuable/highly backed they are, initially. If people spend them quickly, or convert them to bitcoin quickly without collecting interest, they continue to grow in bitcoin value. On the other hand if people hold onto them over time, they become less valuable as the savers get paid for keeping them.
Criticisms: It’s complicated, and the numbers for interest rates and minting fees are pulled out of thin air. Presumably there’s some range that works well, and could be tested by implementing multiple currency specifications side by side.
What kind of smart contracts does it enable?
The single best resource seems to be https://en.bitcoin.it/wiki/Contracts but it only covers a fraction of the possibilities; for example, I corresponded with Hearn about timestamping for time-lock crypto but that hasn’t been added.
Excellent! I had no idea the bitchain included all those extra fields. So much potential!
Yeah; people don’t realize how Bitcoin can do so much. It’s not (just) sending some irreversible transactions around, as useful as it is to replace Paypal/WU. It’s like every financial cryptographic construct ever got mashed up in an academic paper, except it’s not a paper, it’s a real system in use worldwide and the constructs can be actually used by real people for real things. Of course none of them are easy to use, many aren’t even enabled—but I think the situation is like the early Linux kernel: it’s a rare geek who understands why you would want to hack on a kernel, a rarer techie who can compile kernels and mess with kernel settings, and an even rarer techie who can write a kernel module to do something, but the potential and power are there.
You have just created “a rare geek who understands why you would want to hack on a kernel”. These features are screaming out to be used and may facilitate some of the projects that I have in mind (including some plans for small-scale crowdfunded research).
Do you have any other resources on bitcoin to recommend? In particular I a resource that lists definitively which features are enabled and ready to use in current clients? (I hate accidentally building or planning things that rely on features not yet in the mainstream software. It feels like crippling myself when I try to backport.)
No idea. I have a vague understanding that colored coins are a work in progress and may be added at some point in 2013, but I’ve never seen anyone lay out a list of the scripting language elements with the enabled and disabled ones specified. Since this is such an obscure area of Bitcoin, you probably will have to read the source to find out.
Time for me to do an expected value calculation. Is it worth becoming an expert on bitcoin? Not at all implausible. Apart from the intrinsic value of personal interest there seem to be things that the world needs that bitcoin could help with (an actual prediction market that doesn’t suck for example). Some things that the world needs it is possible to create ways to be paid for doing. It aligns well with my aptitudes.
A very tempting rabbit hole to jump down.
a distributed prediction market and distributed exchange with built in escrow would do wonders for making bitcoin an attractive financial instrument.
Creating a bitcoin prediction market is a great idea.
One thing I’ve been wondering recently is why more people don’t create Bitcoin alternatives. Many of the alternatives that already exist have ridiculously high market capitalizations, with what appears to be minimal marketing and almost zero rationale for their existence. If you could create an alternative that had “baked in” solutions to some of the problems people complain about re: long-term adoption of bitcoin (limited divisibility, difficulty of transferring in/out of USD, volatility (somehow fix this? e.g. if the protocol was somehow robust to high-frequency trading, perhaps that would deal with volatility as HFT people traded using volatility pumping strategies), performance problems as the network grows large, general awkwardness of waiting for confirmations and stuff), or had some kind of killer app that bitcoin didn’t have, and then put an actual marketing push behind all this (and bribe the existing exchanges to accept it or find a way to piggyback on the existing bitcoin infrastructure), it seems like it could be a great way to make money.
I’m considering seriously. In the last several days I’ve analyzed the technology already in widespread, looked at the possibilities and explored the limits. The potential is huge. Cryptography is like wizardry that works in the real (abstract) world!
The main not-ridiculously-pointless alternative on the horizon is ripple, which has a different focus to bitcoin. If they have marketing awesomeness behind them they could get people to play their game. (Low probability, high reward for them.)
An awesome cryptographic feature of the bitchain mining system is that new bitchains can be mined using the same computational resources that bitcoin miners use for negligible extra effort. ie. Instead of creating lottery tickets and entering them in one lottery they can create lottery tickets and enter the same ticket in 20 different lotteries. This drastically lowers (but doesn’t remove) the difficulty in getting alternate bitchains to be secure.
Is this a significant problem? 1 satoshi (the minimum division) is 0.000 000 01 BTC. That’s… not very much. And from one I understand changing the system such that it could allow lower divisions would be plausible if it became necessary. (ie. It is the kind of change that could easily be implemented and adopted by a supermajority of computing power of the miners and thereby become “distributed official”.)
The ‘ripple’ currency system has dreams of solving this via webs of (limited) trust and social networks. If this were solved in the ripple network it would also incidentally solve it in any currency that is significantly used.
This volitilty can be solved within bitcoin itself via the adoption of “coloured coins”. Essentially, this means credible institutions can issue “MyBankUSDcoins” for example, with coins with a certain history being declared redeemable at a specified rate via whatever banking mechanism that kind of company does. Similarly pecunix could offer “Pecunix coins” which it backs with gold (its normal business model). Then transactions can be made in gold, or USD, or AUD via the same technology (and distributed security structure) that bitcoin uses.
The above is possible right now. It requires no change to the bitcoin (mining server side) software. It is just a matter of people with sufficient social influence potential making it happen and customers (including merchants) being convinced that they want it.
Yes, systems without the awkwardness would have an advantage.
Has anyone tried to solve the marketing issue by giving their coin a referral system similar to what Paypal had in the early days? I’m not 100% sure about this, but I think maybe the way it worked was if you referred your friend to Paypal, you would get $x and your friend would get $x. Supposedly this was a big reason why Paypal beat out competing services in its early days, ’cause people had this monetary incentive to spread it. (I have an acquaintance who worked for a startup that competed with Paypal and lost; I could probably get in contact with him if anyone has any questions.)
Hm. Sounds like colored coins have all the advantages of Bitcoin, plus lack of volatility. So if they existed then would we ditch regular Bitcoins for them? How much software would need to be changed in order to implement colored coins?
Ultimately, I suspect that Bitcoin is a product that no one really wants that bad, and ultimately the interest in them is mostly from speculation… if there’s any interest in using them as an actual currency, it’s liable to be faked interest from people who have large bitcoin balances and would profit if that happened.
Well there was a lot of interest in Bitcoins in southern Europe following the Cyprus, bank account confiscations. Basically, if people are treating putting money into Bitcoins as an alternative to putting it in banks.
Alas no. They have most of the advantages of bitcoin since they are still bitcoin in full. However, whatever additional value the ‘color’ grants to the coin over the base worth os a BTC is based on trust in a specific agent. They are nothing more than a promise from a supplier to redeem the coins into another form of value at some time in the future. It is a trade off. Convenience and stability purchased by requiring trust.
The regular coins are still what are required to produce coloured coins (the name is really good analogy). Coloured coins being used more would seem to lower the expected price of uncoloured bitcoin, but they couldn’t make uncouloured coins worthless since they themselves entail a demand for them.
It would mean taking the existing ‘beta’ implementations, making them stable and having people start to use them. Merchants accepting them automatically (and being able to tell they are worth more than uncouloured coins) would also require making improvements to whatever software it is that they use now.
I thought merged mining only worked for Bitcoin+Namecoin? I was looking into mining Bitcoin and Litecoin simultaneously, but it looked like it simply wouldn’t work with
cgminerand in general.Merged mining requires a small change to the bitchain of all but one of the currencies. So you can merge mine bitcoin + 20 merge mining enabled currencies OR you can merge mine litecoin + 20 merge mining enabled currencies.
Currencies that can currently be merge mined with BTC (and that I am aware of):
NameCoin—Store aliases and soforth conveniently. Is complementary to bitcoin for some use-cases.
DevCoin --- (Roughly speaking) a charity for funding opensource developers.
TonalCoin—Which somehow is also embedded in the bitcoin blockchain and allows bitcoins to be traded using a different number system.
IxCoin—Utterly pointless attempt to cash in on crypto speculation.
I have been looking into the protocol for how this works. It doesn’t require anything additional to be added (to the bitcoin daemon). To be practical it does require that clients are made that process them conveniently, this is what is the work in progress and does seem to be at the ‘usable in a stable form some time this year’ point.
The above distinction matters to me because it influence whether it is something I can just use if I find customers who want it and I am willing to do some coding to facilitate it versus features that require creation of a Bitcoin Improvement Proposal then the politically influence of a miners to start adopting it. For example zerocoin (the solution to anonymity and privacy) would require huge political advocacy to get into play in the existing bitcoin currency.
Indeed, my research has failed to turn up such a document too. The information that is out there is also significantly out of date unless pieced together rather carefully. A summary of the availability of the features that are of interest. I am reasonably confident in these findings but not certain. (90% confidence.)
Timelocked transactions: The protocol (and actually in use bitchain) supports it but miners will nearly always reject transactions with future dates. This seems stupid to me but there doesn’t seem to be any strong pressure to start accepting them. It is still possible to create such transactions and store them until the time is up then send them in. This solves some timelock usecases but the possibility of double spending prevents it solving many other use cases.
Multisignature transactions. ie. Escrow. Transactions that require n of m signatures in order to spend the output. These work. Technically. You can create them manually via the command line, send them to the blockchain and then use the command line and use of or access to someone who will cooperate with n of m signatures to create transactions that spend them. But current clients don’t do this for you (except in the crippled case whereby if you have m of m keys in your wallet it will deign to let you spend them.) But this feature is there to be used if you want it.
Zero Knowledge Proofs eg. Zerocoins. Reliable privacy and anonymity right there within bitcoin. The research and cryptographic technology is available. As are proof of concept reference implementations. But this system (or an equivalent or superior one) isn’t accepted and likely will not be for at least several years. An alternate currency could do it today.
Done. Now I’ll have to read the source even more to work out how to make clients that use the (available) features to serve my purposes.
Well, good luck. I don’t know a lick of C++ or real crypto coding, so as interesting as I find it all to read and watch and write essays about, don’t expect any help from me. I hope you succeed in getting some of the features in more general use.
But also risks there being some minor problem with the unpublished Zerocoin proposal which de-anonymizes everyone who ever uses the functionality. Since laundries/mixes currently seem to be working pretty well, it’s better to be conservative and get it right, than push for it to be implemented right now and in 2 years, watch every transaction be de-anonymized.
On a related note do you have any thoughts on ripple? It seems to offer huge potential convenience if ‘distributed payment processing’, which could work well to enable bitcoin to be easily spent and traded. From what I can tell their business model is based on making an essentially free, incredibly convenient service and if it takes off it will produce value in their own currency of which they will (allegedly) be distributing 80% for free. I don’t precisely understand how it magically operates as a distributed currency exchange and payment processor but it seems plausible (if enough people want it).
Reading about it right now, I would analogize it to an automated barter system, the hawala money transmission system, pre-modern banking’s system of drafts and correspondents like the Medici Bank, and Szabo’s bitgold minting.
If I’m understanding the wiki documentation right, you can basically think of Ripple as a Bitcoin in which everyone can issue as many bitcoins as they want but each user produces a different kind of bitcoin. The only reason anyone is willing to hold a wedrifid-coin is because they know you in real life and expect you to pay up, or the reason someone is willing to hold gwern-coins is because I have in the past upheld my agreements to, say, order something off Amazon.com for them. I might decide to accept wedrifid-coins and now the wedrifid-coin holders have the option to either shake you down in real life or buy something off Amazon via me; and if someone else decided to accept gwern-coins, a holder of wedrifid-coins could convert to gwern-coins and then convert to someoneelse-coins.… Automated barter of personal IOUs, almost—like those circles you hear of involving scores of people in which A gives something to B, B gives something to C, and so on to Z who gives A what A really wanted (most famously used in kidney transplants; I need a transplant, but my brother is not compatible with me, so he gives his kidney to B, and B’s son gives a kidney to C, and C’s father gives me a compatible kidney).
Since you’re deciding whose coins to accept directly or transitively, Ripple seems to pass the security buck to your decision there. This obviously makes Ripple’s job much much easier. Since you’ll be deciding basically on social or extra-network factors, Ripple also doesn’t bother with any worries about anonymity. (Their list of features re mining, if I’m understanding the architecture right, less than impressive in this light.)
There’s also a centralized aspect which is nasty and probably part of how they plan to monetize it.
So yeah, interesting. I don’t feel compelled to trade in my one bitcoin for any Ripple stuff, though.
I think it operates pretty much the same way a PGP web of trust operates. You specify a few public keys who you trust, and you can bootstrap from there.
So (I assume) the way it can operate as a distributed currency exchange is if there is someone with Yen out there and someone with USD out there and either Yen-guy or USD-guy trust each other directly or their is a web of trust between them. So RandomJoe can sell Yen-guy coins for stuff and buy USD-coins then USD guy will give actual cash to him.
Yes, I think that’s how it could go down. Or Yen-guy and USD-guy could exchange currency directly, or they could do it hawala-style indirectly (for example: Yen-guy could hand out yen to anyone wandering into his Tokyo shop with USD-guy coins, and then USD-guy, to settle his bills with Yen-guy, can fly in and visit Yen-guy with a case of wine which Yen-guy can drink or sell).
I wonder if a form of coin-locking could be used to create simulated versions of other currencies. Depending on the demand and level of trust in the given currency, a given bitcoin sum could be locked in such a way that it gets automatically distributed to all holders at some future point (if they don’t choose to hold onto the simulated currency instead) in proportion to how much they have. The choice to hold or drop the simulated currency would be a way of measuring its value against the bitcoin.
I can think of various ways to facilitate this using the techniques I am considering (or considered and discarded) for the creation of a distributed prediction market. The details of the precise implementation would depend on precisely what your goal was. I can’t fully determine what you are aiming at from what you have said so far? More detail? A use case?
What I have in mind is something a little bit like the zerocoin proposal, in the respect that it’s an alternate coin which can be translated to Bitcoin and back as part of how the system works.
But instead of being fixed with respect to the bitcoin it would increase or decrease in relative value in response to market signals. There would be an exchange rate determined by dividing the number of spendable units of alternate currency against the number of bitcoins locked behind it.
So what you need is a series of rules for controlling that number that tend towards economic stability. Which is itself kind of tricky. I had thought locking the coins by a time-lock could do it, but I haven’t thought of a way to base an exchange rate on that. Empirical work with simulator or a smart math person with a whiteboard might turn up something.
The following uses incentives to accomplish something similar-in-principle:
Alice creates 100 dollarcoins at a cost of 1 bitcoin. She transfers 1 to Bob, he can turn it into 0.01 bitcoin.
The dollarcoin has a specification that includes things like interest payments. If Bob waits 30 days after receiving a dollarcoin he gets 1% of its nominal bitcoin value in interest. He gets another 1% payment 30 days later, and so forth. But if he spends it right away or converts it to bitcoin there are no payments.
Claire decides to use the dollarcoin specification to mint some dollarcoins. She has to pay a 1% minting fee, so 1.01 bitcoins to make 100 dollarcoins. The backing rate becomes 200 dollarcoins to 2.01 bitcoins.
Thus the more dollarcoins get minted the more valuable/highly backed they are, initially. If people spend them quickly, or convert them to bitcoin quickly without collecting interest, they continue to grow in bitcoin value. On the other hand if people hold onto them over time, they become less valuable as the savers get paid for keeping them.
Criticisms: It’s complicated, and the numbers for interest rates and minting fees are pulled out of thin air. Presumably there’s some range that works well, and could be tested by implementing multiple currency specifications side by side.