Except that the ability to solve a simple puzzle may be different from the ability to recognize spam (which is what we are really trying to stop here). Even if you had a computer that was better at solving CAPTCHAs than humans are, you might still be unable to detect the specific class of unwanted behavior that you were trying to defend against. Now, if the CAPTCHA was asking you to label a series of short messages "spam" or "not spam," then perhaps your point would hold...except that it would be far too annoying for most people to deal with.
There must still be computational areas in the visual domain where we humans are way more efficient.
Even if that is the case, there is still a relatively straightforward attack on captchas: the mafia porn site. It is generally easier to use a mechanical turk to decode captchas than to attack captchas algorithmically.
The problem is not creating things which are hard for computers to decode, it is creating things which are hard for computers to decode but easy for humans. That is why captchas will ultimately fail: they rely on the idea that there is something that human brains can understand which computers cannot decode, but which computers can still generate.
Except that the US is not ordering any drone strikes against programmers whose software is used on pornography websites. The US may have slipped from its founders' ideals, but we are not quite at the level of Iran.
I am quick to criticize the US government for drifting away from protecting our rights and freedoms, but there is really no comparison with a country like Iran. In the US, when the government considers establishing a national firewall, the citizens criticize the government without fear; in Iran, protesting the firewall that is already in existence can result in being arrested or beaten up by the police. The fact that we can even compare the US to an enemy like Iran, without worrying about angering the censors, shows you just how many freedoms we still have in this country.
He also did not cause any real harm. I guess how far to the left or right one leans determines whether or not the line should be drawn at "causing harm" or "had no business doing it."
If connecting GMRS and MURS stations to the phone system were legal, this would be an easier problem to solve. People in rural areas could set up wireless services for themselves, essentially creating their own cooperative cell phone networks.
Not that the telcos want to change those regulations.
Sorry, I don't get this at all. How would a centralized currency survive if it's heavily utilized for ends that are against the dominant power's policy?
The same way it does right now, with government issued currency? All the black market activity with Bitcoin combined doesn't even come close to the number of dollars used in black market transactions each year, let alone pesos, yuan, etc. The reason digital cash can be used in black market transactions -- the reason that black market transactions benefit from it -- is that it can be spent and received anonymously.
The problem is analogous because both systems require a trail of all transactions
Actually, I have to retract this point, as someone else has already noted that Chaum's result does not necessarily apply to Bitcoin: Bitcoin does not allow for offline transactions. Chaum's result, as far as I know, applied to any digital cash system in which offline transactions are possible, so perhaps Bitcoin trades offline transactions for the ability to be scalable without a central authority.
Not that I think this really helps Bitcoin's case, since offline transactions are important for digital cash (at least in my opinion). I want to be able to spend my digital cash even when an Internet connection is unavailable, expensive, etc. Gas pumps, soda machines, charity boxes, public transit buses and so forth should not have to connect to some P2P network just to accept a small payment, and digital cash should be easy to deploy in less developed regions where Internet access may be scarce. Again, if a digital cash system that allowed offline transactions were to be deployed, it would quickly usurp Bitcoin.
Since you've been telling a better cash is possible with Chaum's scheme, I'd like to remind you that it is actually a 90's era thing, and there are already a multitude of such systems in place.
Keep in mind that I did not specifically say that Chaum's scheme should be deployed, I was only referring to the result on digital cash scalability. Also, what difference does it make if it was developed in the 90s -- RSA was developed in the 70s and it is still widely used.
I am aware that there have been efforts to deploy digital cash, but unfortunately these efforts have not been on a large enough scale. We need something like Berkshares, but with digital cash, if not something on an even grander scale:
Way to not read what I wrote. What I said was that people accept government issued currencies because their governments will not accept tax payments made with anything else, and paying taxes is mandatory.
It's OK though, you can ignore what people say and continue to live in some fantasy world where demand has nothing to do with the value of money, and where people just magically start thinking that currencies are useful.
The only dishonesty that has ever been involved with drug prohibition laws has been on the part of lobbyists, politicians, and the police. Drugs are not made illegal to protect people from drug users, despite the early drug prohibition claims about black cocaine users becoming unstoppable monsters that could survive a bullet to through the heart. Drug prohibition is about increasing police power, increasing corporate profits, and attacking civil liberties.
the down-the road consequences of losing a paramilitary force that kills with impunity, siezes civilian property to fund its operations, operates inside of and outside of US borders, has more signals intelligence capability than the intelligence services of most nations, and which both creates and enforces drug possession laws without any democratic process.
How much suffering is the DEA willing to inflict for the continued militarization and increase in power of American law enforcement agencies, and the continued expansion of executive branch power at the expense of civil liberties and democracy?
Well in that case, I have nothing to argue, I was under the impression that your claim was that Bitcoin would not suffice even in that case. (Quote: "In which case it would only take a digital currency system that does not suffer from Bitcoin's inherent scalability problem").
I did not claim that the black market would not / could not make use of a digital cash system; it is that our focus should not be on black markets. Yes, black markets do benefit from digital cash, regardless of whether the cash is Bitcoin-like or issued by a central authority. If Bitcoin were relegated to black market use, it would still die when a better digital cash system was deployed. The point here is that if the only advantage Bitcoin has is that it is a digital cash system (which is the only reason black market transactions have even occurred with Bitcoin, and even then, only because Bitcoin can eventually be exchanged for some national currency), then a better system will come along at some point and Bitcoin will quickly be dropped in favor of that system.
you see Bitcoin as a digital cash in Chaum's terms. Technically it isn't.
Then explain the difference, because neither the original Bitcoin paper mentions Chaum's work (or any previous work on digital cash) nor does the Bitcoin Wiki's page on scalability. People keep saying that Chaum's result does not apply to Bitcoin, but nobody seems to be able to say why that is the case.
You need an outside threat to prevent a centralized authority from running away with your digital cash
Or we can have an insurance system like the FDIC, or even better, have the same government or bank that issues a nation's currency issue digital cash (in which case there is no "running away" to speak of).
That's the whole reason for the concepts of predefined money supply and proof-of-work.
Except that the predefined money supply is a terrible idea for economic reasons.
You either didn't read what I said or didn't understand it. I sort of explained it from the implementation side, so let me try again from the "result" side.
You are not understanding what I am saying: it is a general result for digital cash systems that the amount of information that must be stored or transmitted will grow with the number of transactions. It does not matter how you manipulate the token, represent the token, or allow the token to be used in combination with other tokens; the information has to be stored somewhere. The fact that Bitcoin keeps pruning the tokens does not mean that it escapes this result; the number of bits left over after pruning will increase over time (or else Bitcoin violates some other assumption about privacy or protection from double spending, or there is actually a hidden money issuing authority in Bitcoin that can renew the money).
Unless you would like to explain why Chaum's result does not actually apply to Bitcoin; the original Bitcoin paper does not appear to mention Chaum's work at all (or any previous work on digital cash, which is pretty troubling).
I would rather not be at the mercy of Microsoft, Apple, Google, or Facebook. They do not need to change their API, they can just change the licensing and suddenly my software is threatened.
Except that with an open source project, you can always fork -- and if an API change is so drastic that an entire software ecosystem is threatened by it, a fork is likely to happen (or a project may simply maintain two versions -- Apache does this). Firefox has come pretty close, but extension developers do not represent a large enough ecosystem for the community to fork Firefox, and the API changes are not drastic enough to necessitate such a thing.
Open APIs are a way to say, "Proprietary software platforms that you can write your own software for!" The word "trap" does not even come close to describing the nature of these systems, which are attempts by certain companies to control large software ecosystems (e.g. what happened with Windows).
Are we still talking about black/gray markets here?
No, because black/grey markets are not the primary reason for deploying digital cash. Digital cash is advantageous because it adds security to electronic transactions, not because it enables people to skirt laws or regulations.
It/is/ a necessity. See: e-gold, Central issuer is a single point of failure. It allows regulators direct access to the economy. Not helpful in the gray market case.
Central payment processing is a point of failure; central issuance of money is common and I would argue necessary for a currency to be successful. Note that e-gold was like Paypal in that you needed to be online and logged in to process a payment, this is not a digital cash system (at least not in the cryptographic sense of the term). Digital cash systems should allow transactions to take place offline, which in this context means without having to connect to any central processing authority to complete the transaction.
Again, gray market transactions are not the goal of digital cash systems. The point of digital cash is to allow secure electronic payments, which means not only protecting against eavesdroppers, but also against unscrupulous merchants who might try to get more money than the customer agreed to and against unscrupulous spenders who might try to commit various forms of fraud. A secondary goal is anonymous transactions, which allow the spender's privacy to be preserved from other parties in the system. That digital cash can enable online black/gray markets is irrelevant, except in that it creates a new challenge for law enforcement.
Additionally, regulation is not a bad thing, except perhaps in the minds of ardent libertarians and anarchists. The goal in digital cash should not be to defeat regulation; we large corporations and especially banks to be regulated. The goal, once again, is to protect people from various forms of fraud and theft.
I'm somewhat familiar with the paper, I hadn't thought it was directly applicable to the case of Bitcoin (isn't it mostly about Chaum's own scheme?), so I'll have to read it again.
My understanding of Chaum's result is that it is generic and that it applies to any digital cash system that allows offline transactions and token reuse, and that the result basically says that if you have security and privacy then the size of a token must grow with the number of transactions. This issue has been avoided in some systems by distributing the storage of the token along each node in the transaction chain, the "receipts" system, although if I remember correctly the receipts must eventually be sent to the issuing authority.
Though basically, in Bitcoin's case, keys can be and are merged, so there is the technical foundation to make shrinking possible (isn't token renewal in Chaum's terminology analogous to key merging?).
I believe that "renewal" in Chaum's case referred to a transaction with the issuing authority itself, where the authority would receive a token that had been used in many transactions and the spender would receive a fresh token of equal value. Clearly this is not possible in Bitcoin's case, and the "merging" process should not be able to evade Chaum's general result on tokens becoming larger. Again, it has been a while since I read that paper, so I could be wrong in my understanding of it.
The blockchain will still undergo a lot of growth if more people begin using Bitcoin (since the currency base will be divided between more wallets, and there will be greater numbers of recent transactions that are not yet pruned), but it doesn't geometrically expand forever. When Bitcoin stabilizes at a certain level of market saturation the blockchain will also stabilize in size.
Which implies that Bitcoin is not secure. Somewhere, the storage or bandwidth require must grow with the number of transactions that any token is used in in any secure digital cash system, regardless of how the token is manipulated. If that is not the case, then either the system does not provide anonymity, or it does not provide security against doubt spending (I suspect the former in the case of Bitcoin).
I want to know that the central bank won't just start issuing more currency and devalue mine.
That's nice, but if your money supply cannot grow then your economy is going to have problems. Claiming that you can keep dividing money into smaller units does not solve this problem.
I want to send money to Wikileaks, but all the banks are forbidding it.
That is the point of digital cash. What, did you think that Bitcoin was the be-all and end-all of digital cash systems, or that you could not possibly have a distributed digital payment system with centralized currency authorities?
I want to pay someone over the internet without Mastercard or Paypal or whoever taking a 2% cut.
See above.
There is no other currency that currently has ALL of those advantages of Bitcoin
The only advantage that is not shared with other digital cash systems is the "banks cannot cause hyperinflation" point, and the Bitcoin approach to this is to force a deflationary currency, which is even worse. How would you like to try repaying a loan (and yes, sometimes people need to take out a loan) with a deflationary currency? How would you like a world where people hoard the currency because it is profitable to do so?
You should take the time to read some of the work that was done on digital cash systems over the past 25 years before making statements about how great Bitcoin is. Yes, Bitcoin fits in nicely with the anarchist dream of economies and money without government, but in the real world we need to pay our taxes and we need courts to decide how to settle monetary disputes. We should be working to deploy digital cash systems that allow us to continue to do such things, while having the advantages of digital cash: secure online payments, secure offline payments, anonymous payments, etc.
This is a technique that has also been discussed in the literature, called "receipts." That is still not scalable, since the amount of disc space required to run the system at all will continue to grow over time; worse, if the Bitcoin economy is growing, the rate at which disc space is consumed will accelerate. The receipts approach is generally only useful if you can delete the log -- for example, if the token is renewed by an issuing authority, which is not applicable in Bitcoin's case.
Wrong, people demand government issued currencies because they need it in order to pay taxes and other debts; failure to make such payments can result in the loss of one's property or freedom. There is a legal structure surrounding government issued money, which is where demand for that money comes from: courts determine property values and damages in terms of he government's currency, tax collectors will only accept payments in government currency, banks will only accept debt payments in government currency, etc. The problem with the 10th-grade "people believe their money can be used for xyz" explanation is that it completely avoids any discussion of why people use their government's currency in the first place, when they could use some other currency like Bitcoin or just resort to barter (which people do resort to, and which they still wind up having to pay taxes on).
Why do the currencies of ineffective or defunct governments stop being valuable? Why do people stop believing that the money can be used to buy goods and services? Answer that question, and you will see why Bitcoin will not work in the long run.
A design that is resilient to regulation is needed in order to have the advantage in the first place,
Actually, the biggest advantage of digital cash is that it prevents an untrustworthy merchant from raiding customer accounts (e.g. if a merchant is compromised, credit card data could be used to make unauthorized payments). Anonymity is a secondary goal, and even anonymous payments do not free the parties in the system from regulation -- there is plenty of regulation on cash transactions, which offer a similar level of anonymity.
at least some sort of decentralization is practically a requirement
Yes, that is true -- the payments should be decentralized in a digital cash system. It is not necessary or advantageous for currency units to be created in a decentralized way (like Bitcoin), and it actually presents disadvantages, both technical and economic. Other digital cash systems are more like paper money: issued by a central authority, spent in a decentralized manner, and periodically renewed (old paper currency is destroyed, new paper currency is issued, and likewise old digital cash tokens are replaced with fresh tokens of equal value in many digital cash systems).
If Bitcoin is valuable enough as a currency, and scalability issues become a problem, the network itself would inevitably become the transaction backbone of a most wider structure, comprising many interlinked central issuing authorities
Except that Bitcoin does not allow for central issuing authorities, and there is no advantage in having such authorities use Bitcoin -- they are inherently trusted, and it is reasonable to assume that they will trust each other (or at least that they trust each other enough to perform more standard transactions -- the real point of digital cash is for day-to-day transactions, not transactions between large institutions).
Third scenario is where we find an ultimate solution to the scalability problem. There are many methods that would make do for the foreseeable future, but there is none that does away with it once and for all. Though, if it all boils down to the renewal of tokens as you mentioned, I think it can still be done with a distributed system.
Not without a trusted party that issues the tokens. This has been discussed by researchers:
Sorry if you don't have SprinerLink access; a non-Springer version might be available somewhere, the paper title is Transferred Cash Grows in Size, author is David Chaum.
The solutions to this involve doing things like storing receipts on hard drives (which only distributes the load, similar to Bitcoin's approach) and token renewal by an issuing authority. Token renewal is not hard in a world where Internet access is widely available and should probably be our preferred approach.
Slashdot Mirror
Except that the ability to solve a simple puzzle may be different from the ability to recognize spam (which is what we are really trying to stop here). Even if you had a computer that was better at solving CAPTCHAs than humans are, you might still be unable to detect the specific class of unwanted behavior that you were trying to defend against. Now, if the CAPTCHA was asking you to label a series of short messages "spam" or "not spam," then perhaps your point would hold...except that it would be far too annoying for most people to deal with.
In which of those cases is the government not interfering?
https://en.wikipedia.org/wiki/Cold_war
https://en.wikipedia.org/wiki/SOPA
http://www.wired.com/threatlevel/2010/07/ticketmaster/
There must still be computational areas in the visual domain where we humans are way more efficient.
Even if that is the case, there is still a relatively straightforward attack on captchas: the mafia porn site. It is generally easier to use a mechanical turk to decode captchas than to attack captchas algorithmically.
"Anything a computer can generate it can understand."
Thus explaining the prevalence of these:
https://en.wikipedia.org/wiki/Ciphertext_only_attack
The problem is not creating things which are hard for computers to decode, it is creating things which are hard for computers to decode but easy for humans. That is why captchas will ultimately fail: they rely on the idea that there is something that human brains can understand which computers cannot decode, but which computers can still generate.
As opposed to Democrats? Some of us want to kick both parties out of our government.
Except that the US is not ordering any drone strikes against programmers whose software is used on pornography websites. The US may have slipped from its founders' ideals, but we are not quite at the level of Iran.
I am quick to criticize the US government for drifting away from protecting our rights and freedoms, but there is really no comparison with a country like Iran. In the US, when the government considers establishing a national firewall, the citizens criticize the government without fear; in Iran, protesting the firewall that is already in existence can result in being arrested or beaten up by the police. The fact that we can even compare the US to an enemy like Iran, without worrying about angering the censors, shows you just how many freedoms we still have in this country.
He also did not cause any real harm. I guess how far to the left or right one leans determines whether or not the line should be drawn at "causing harm" or "had no business doing it."
If connecting GMRS and MURS stations to the phone system were legal, this would be an easier problem to solve. People in rural areas could set up wireless services for themselves, essentially creating their own cooperative cell phone networks.
Not that the telcos want to change those regulations.
Sorry, I don't get this at all. How would a centralized currency survive if it's heavily utilized for ends that are against the dominant power's policy?
The same way it does right now, with government issued currency? All the black market activity with Bitcoin combined doesn't even come close to the number of dollars used in black market transactions each year, let alone pesos, yuan, etc. The reason digital cash can be used in black market transactions -- the reason that black market transactions benefit from it -- is that it can be spent and received anonymously.
The problem is analogous because both systems require a trail of all transactions
Actually, I have to retract this point, as someone else has already noted that Chaum's result does not necessarily apply to Bitcoin: Bitcoin does not allow for offline transactions. Chaum's result, as far as I know, applied to any digital cash system in which offline transactions are possible, so perhaps Bitcoin trades offline transactions for the ability to be scalable without a central authority.
Not that I think this really helps Bitcoin's case, since offline transactions are important for digital cash (at least in my opinion). I want to be able to spend my digital cash even when an Internet connection is unavailable, expensive, etc. Gas pumps, soda machines, charity boxes, public transit buses and so forth should not have to connect to some P2P network just to accept a small payment, and digital cash should be easy to deploy in less developed regions where Internet access may be scarce. Again, if a digital cash system that allowed offline transactions were to be deployed, it would quickly usurp Bitcoin.
Since you've been telling a better cash is possible with Chaum's scheme, I'd like to remind you that it is actually a 90's era thing, and there are already a multitude of such systems in place.
Keep in mind that I did not specifically say that Chaum's scheme should be deployed, I was only referring to the result on digital cash scalability. Also, what difference does it make if it was developed in the 90s -- RSA was developed in the 70s and it is still widely used.
I am aware that there have been efforts to deploy digital cash, but unfortunately these efforts have not been on a large enough scale. We need something like Berkshares, but with digital cash, if not something on an even grander scale:
https://en.wikipedia.org/wiki/Berkshares
I am not aware of any such effort with digital cash, but perhaps you are?
the arguments against it on these grounds are actually similar to the ideas against Esperanto.
Which is even more obscure than Bitcoin despite having been developed and promoted for over a century.
Way to not read what I wrote. What I said was that people accept government issued currencies because their governments will not accept tax payments made with anything else, and paying taxes is mandatory.
It's OK though, you can ignore what people say and continue to live in some fantasy world where demand has nothing to do with the value of money, and where people just magically start thinking that currencies are useful.
The only dishonesty that has ever been involved with drug prohibition laws has been on the part of lobbyists, politicians, and the police. Drugs are not made illegal to protect people from drug users, despite the early drug prohibition claims about black cocaine users becoming unstoppable monsters that could survive a bullet to through the heart. Drug prohibition is about increasing police power, increasing corporate profits, and attacking civil liberties.
the down-the road consequences of losing a paramilitary force that kills with impunity, siezes civilian property to fund its operations, operates inside of and outside of US borders, has more signals intelligence capability than the intelligence services of most nations, and which both creates and enforces drug possession laws without any democratic process.
FTFY
How much suffering is the DEA willing to inflict for the continued militarization and increase in power of American law enforcement agencies, and the continued expansion of executive branch power at the expense of civil liberties and democracy?
FTFY
The DEA does do a lot of important things
I know, right..
Yes, this is one agency that America really needs to keep around.
Well in that case, I have nothing to argue, I was under the impression that your claim was that Bitcoin would not suffice even in that case. (Quote: "In which case it would only take a digital currency system that does not suffer from Bitcoin's inherent scalability problem").
I did not claim that the black market would not / could not make use of a digital cash system; it is that our focus should not be on black markets. Yes, black markets do benefit from digital cash, regardless of whether the cash is Bitcoin-like or issued by a central authority. If Bitcoin were relegated to black market use, it would still die when a better digital cash system was deployed. The point here is that if the only advantage Bitcoin has is that it is a digital cash system (which is the only reason black market transactions have even occurred with Bitcoin, and even then, only because Bitcoin can eventually be exchanged for some national currency), then a better system will come along at some point and Bitcoin will quickly be dropped in favor of that system.
you see Bitcoin as a digital cash in Chaum's terms. Technically it isn't.
Then explain the difference, because neither the original Bitcoin paper mentions Chaum's work (or any previous work on digital cash) nor does the Bitcoin Wiki's page on scalability. People keep saying that Chaum's result does not apply to Bitcoin, but nobody seems to be able to say why that is the case.
You need an outside threat to prevent a centralized authority from running away with your digital cash
Or we can have an insurance system like the FDIC, or even better, have the same government or bank that issues a nation's currency issue digital cash (in which case there is no "running away" to speak of).
That's the whole reason for the concepts of predefined money supply and proof-of-work.
Except that the predefined money supply is a terrible idea for economic reasons.
You either didn't read what I said or didn't understand it. I sort of explained it from the implementation side, so let me try again from the "result" side.
You are not understanding what I am saying: it is a general result for digital cash systems that the amount of information that must be stored or transmitted will grow with the number of transactions. It does not matter how you manipulate the token, represent the token, or allow the token to be used in combination with other tokens; the information has to be stored somewhere. The fact that Bitcoin keeps pruning the tokens does not mean that it escapes this result; the number of bits left over after pruning will increase over time (or else Bitcoin violates some other assumption about privacy or protection from double spending, or there is actually a hidden money issuing authority in Bitcoin that can renew the money).
Unless you would like to explain why Chaum's result does not actually apply to Bitcoin; the original Bitcoin paper does not appear to mention Chaum's work at all (or any previous work on digital cash, which is pretty troubling).
I would rather not be at the mercy of Microsoft, Apple, Google, or Facebook. They do not need to change their API, they can just change the licensing and suddenly my software is threatened.
Except that with an open source project, you can always fork -- and if an API change is so drastic that an entire software ecosystem is threatened by it, a fork is likely to happen (or a project may simply maintain two versions -- Apache does this). Firefox has come pretty close, but extension developers do not represent a large enough ecosystem for the community to fork Firefox, and the API changes are not drastic enough to necessitate such a thing.
Open APIs are a way to say, "Proprietary software platforms that you can write your own software for!" The word "trap" does not even come close to describing the nature of these systems, which are attempts by certain companies to control large software ecosystems (e.g. what happened with Windows).
Are we still talking about black/gray markets here?
No, because black/grey markets are not the primary reason for deploying digital cash. Digital cash is advantageous because it adds security to electronic transactions, not because it enables people to skirt laws or regulations.
It /is/ a necessity. See: e-gold, Central issuer is a single point of failure. It allows regulators direct access to the economy. Not helpful in the gray market case.
Central payment processing is a point of failure; central issuance of money is common and I would argue necessary for a currency to be successful. Note that e-gold was like Paypal in that you needed to be online and logged in to process a payment, this is not a digital cash system (at least not in the cryptographic sense of the term). Digital cash systems should allow transactions to take place offline, which in this context means without having to connect to any central processing authority to complete the transaction.
Again, gray market transactions are not the goal of digital cash systems. The point of digital cash is to allow secure electronic payments, which means not only protecting against eavesdroppers, but also against unscrupulous merchants who might try to get more money than the customer agreed to and against unscrupulous spenders who might try to commit various forms of fraud. A secondary goal is anonymous transactions, which allow the spender's privacy to be preserved from other parties in the system. That digital cash can enable online black/gray markets is irrelevant, except in that it creates a new challenge for law enforcement.
Additionally, regulation is not a bad thing, except perhaps in the minds of ardent libertarians and anarchists. The goal in digital cash should not be to defeat regulation; we large corporations and especially banks to be regulated. The goal, once again, is to protect people from various forms of fraud and theft.
I'm somewhat familiar with the paper, I hadn't thought it was directly applicable to the case of Bitcoin (isn't it mostly about Chaum's own scheme?), so I'll have to read it again.
My understanding of Chaum's result is that it is generic and that it applies to any digital cash system that allows offline transactions and token reuse, and that the result basically says that if you have security and privacy then the size of a token must grow with the number of transactions. This issue has been avoided in some systems by distributing the storage of the token along each node in the transaction chain, the "receipts" system, although if I remember correctly the receipts must eventually be sent to the issuing authority.
Though basically, in Bitcoin's case, keys can be and are merged, so there is the technical foundation to make shrinking possible (isn't token renewal in Chaum's terminology analogous to key merging?).
I believe that "renewal" in Chaum's case referred to a transaction with the issuing authority itself, where the authority would receive a token that had been used in many transactions and the spender would receive a fresh token of equal value. Clearly this is not possible in Bitcoin's case, and the "merging" process should not be able to evade Chaum's general result on tokens becoming larger. Again, it has been a while since I read that paper, so I could be wrong in my understanding of it.
The blockchain will still undergo a lot of growth if more people begin using Bitcoin (since the currency base will be divided between more wallets, and there will be greater numbers of recent transactions that are not yet pruned), but it doesn't geometrically expand forever. When Bitcoin stabilizes at a certain level of market saturation the blockchain will also stabilize in size.
Which implies that Bitcoin is not secure. Somewhere, the storage or bandwidth require must grow with the number of transactions that any token is used in in any secure digital cash system, regardless of how the token is manipulated. If that is not the case, then either the system does not provide anonymity, or it does not provide security against doubt spending (I suspect the former in the case of Bitcoin).
I want to know that the central bank won't just start issuing more currency and devalue mine.
That's nice, but if your money supply cannot grow then your economy is going to have problems. Claiming that you can keep dividing money into smaller units does not solve this problem.
I want to send money to Wikileaks, but all the banks are forbidding it.
That is the point of digital cash. What, did you think that Bitcoin was the be-all and end-all of digital cash systems, or that you could not possibly have a distributed digital payment system with centralized currency authorities?
I want to pay someone over the internet without Mastercard or Paypal or whoever taking a 2% cut.
See above.
There is no other currency that currently has ALL of those advantages of Bitcoin
The only advantage that is not shared with other digital cash systems is the "banks cannot cause hyperinflation" point, and the Bitcoin approach to this is to force a deflationary currency, which is even worse. How would you like to try repaying a loan (and yes, sometimes people need to take out a loan) with a deflationary currency? How would you like a world where people hoard the currency because it is profitable to do so?
You should take the time to read some of the work that was done on digital cash systems over the past 25 years before making statements about how great Bitcoin is. Yes, Bitcoin fits in nicely with the anarchist dream of economies and money without government, but in the real world we need to pay our taxes and we need courts to decide how to settle monetary disputes. We should be working to deploy digital cash systems that allow us to continue to do such things, while having the advantages of digital cash: secure online payments, secure offline payments, anonymous payments, etc.
This is a technique that has also been discussed in the literature, called "receipts." That is still not scalable, since the amount of disc space required to run the system at all will continue to grow over time; worse, if the Bitcoin economy is growing, the rate at which disc space is consumed will accelerate. The receipts approach is generally only useful if you can delete the log -- for example, if the token is renewed by an issuing authority, which is not applicable in Bitcoin's case.
Wrong, people demand government issued currencies because they need it in order to pay taxes and other debts; failure to make such payments can result in the loss of one's property or freedom. There is a legal structure surrounding government issued money, which is where demand for that money comes from: courts determine property values and damages in terms of he government's currency, tax collectors will only accept payments in government currency, banks will only accept debt payments in government currency, etc. The problem with the 10th-grade "people believe their money can be used for xyz" explanation is that it completely avoids any discussion of why people use their government's currency in the first place, when they could use some other currency like Bitcoin or just resort to barter (which people do resort to, and which they still wind up having to pay taxes on).
Why do the currencies of ineffective or defunct governments stop being valuable? Why do people stop believing that the money can be used to buy goods and services? Answer that question, and you will see why Bitcoin will not work in the long run.
A design that is resilient to regulation is needed in order to have the advantage in the first place,
Actually, the biggest advantage of digital cash is that it prevents an untrustworthy merchant from raiding customer accounts (e.g. if a merchant is compromised, credit card data could be used to make unauthorized payments). Anonymity is a secondary goal, and even anonymous payments do not free the parties in the system from regulation -- there is plenty of regulation on cash transactions, which offer a similar level of anonymity.
at least some sort of decentralization is practically a requirement
Yes, that is true -- the payments should be decentralized in a digital cash system. It is not necessary or advantageous for currency units to be created in a decentralized way (like Bitcoin), and it actually presents disadvantages, both technical and economic. Other digital cash systems are more like paper money: issued by a central authority, spent in a decentralized manner, and periodically renewed (old paper currency is destroyed, new paper currency is issued, and likewise old digital cash tokens are replaced with fresh tokens of equal value in many digital cash systems).
If Bitcoin is valuable enough as a currency, and scalability issues become a problem, the network itself would inevitably become the transaction backbone of a most wider structure, comprising many interlinked central issuing authorities
Except that Bitcoin does not allow for central issuing authorities, and there is no advantage in having such authorities use Bitcoin -- they are inherently trusted, and it is reasonable to assume that they will trust each other (or at least that they trust each other enough to perform more standard transactions -- the real point of digital cash is for day-to-day transactions, not transactions between large institutions).
Third scenario is where we find an ultimate solution to the scalability problem. There are many methods that would make do for the foreseeable future, but there is none that does away with it once and for all. Though, if it all boils down to the renewal of tokens as you mentioned, I think it can still be done with a distributed system.
Not without a trusted party that issues the tokens. This has been discussed by researchers:
http://www.springerlink.com/content/ew2cm5yd0kg797yf/
Sorry if you don't have SprinerLink access; a non-Springer version might be available somewhere, the paper title is Transferred Cash Grows in Size, author is David Chaum.
The solutions to this involve doing things like storing receipts on hard drives (which only distributes the load, similar to Bitcoin's approach) and token renewal by an issuing authority. Token renewal is not hard in a world where Internet access is widely available and should probably be our preferred approach.