$7 today?!!

[SgtSpike]

But see, as long as banks are used, they can manipulate the money supply.  Fractional reserve, and all of that.  If people accept a bank balance as good as coins themselves, then banks can lend out more BTC than they actually have, effectively inflating the BTC supply.  And just as easily, they can withdraw their lending to shore up their reserves and deflate the BTC supply.

If you don't believe banking will be forced, but rather optional, then face-to-face transactioning needs to be solved before we get to that level of Bitcoin usage.  As it is now, you can use a mobile phone to send payments, but again, it can't be trusted at 0-conf, or people will find ways to easily trick the retailers.  Aside from green addresses, which would be controlled by banks or bank-like institutions, or bank-account-to-bank-account balance transfers, I just don't see a way to make point of sale payments with Bitcoins.  Please do prove me wrong on that front though.

IMO, any transaction for a retailer could be trusted at 1-conf, but no one is going to wait 10 minutes for their payment to go through, which makes it just as useless for POS as 6-conf.

[1714891141]

1714891141

[1714891141]

1714891141

[1714891141]

1714891141

[1714891141]

1714891141

[1714891141]

1714891141

[tvbcof]

...
There remains room here for traditional 3rd-party banking services, but it is still a matter of choice rather than force as SgtSpike pointed out.

I run a full client, but also use Instawallet.org for times when it makes sense and keep some spending money there.

A Bitcoin banking service which allows me alone to control my private keys is a giant step ahead of traditional banks as we know them, but...

The real magic of Bitcoin that attracted me in the first place is that the entire economy can be operated independently and autonomously 'normal' people with very little organisation.  When operation of the network requires specialized connectivity available mainly within datecenters or large private organizations, this will spell to me a very significant weakness against attack in the system and thus a much less interesting solution.

[casascius]

But see, as long as banks are used, they can manipulate the money supply.  Fractional reserve, and all of that.  If people accept a bank balance as good as coins themselves, then banks can lend out more BTC than they actually have, effectively inflating the BTC supply.  And just as easily, they can withdraw their lending to shore up their reserves and deflate the BTC supply.

I think Bitcoin banks could very easily and credibly prove they have the BTC they say they have via the digital signature function.

The only slightly more difficult part would be them credibly proving that the amount they have on deposit is accurate.  Not impossible though.  I propose a simple straightforward way.

One way they could do this is to create an anonymous list of all their depositors and put them in some sequence.

Example, MtGox could say and sign, "we have 1000000 BTC on deposit", and create an ordered list of balances.  Each entry on the list would have a start and end range, for example, the first entry for example 25 BTC would "start" at 0 and "end" at 25, and the second entry for example 100 BTC would "start" at 25 and "end" at 125.  Each entry would have a public key published with it, and the entire list would be signed by MtGox.  The last item on the list would "end" at 1000000 BTC.

Then MtGox would give each depositor a signed statement that "your deposit, acct xyxy is on our list at <start> and <end>", along with the matching private key for that entry.

Two depositors who could produce messages both claiming the same entry as their deposit could publicly prove that the bank misrepresented its liabilities.

The private and public key scheme included with each line would simply allow the individual depositors to collaborate via some independent registry, so they could anonymously claim ownership of their funds in an effort to look for double allocations without sharing their account numbers or identities at large.  (e.g. they would each sign messages saying "These funds are mine, contact me <how> if you think they're yours", where <how> is generally going to be "publish your contact details encrypted with <public key> and proof that the bank signed a message saying these are yours, and I'll find you.")  While not every depositor would participate, it would only take one collision to discredit the bank - which would be a strong incentive to keep the bank honest.

[mb300sd]

Multisig transactions should solve all the banking problems. Your deposits are with them, but they cannot move them unless you provide the other signature. Double spends can be prevented because the bank will not sign a double spend. The "bank" is basically reduced to a trusted third party that prevents double spends, allowing merchants to trust 0-conf transactions.

[DeathAndTaxes]

Yeah, exactly waspoza.  A merchant would be much more likely to trust a transaction coming directly from a bank than from a consumer standing there with a handheld Bitcoin client in their hand.

The next step, of course, would be bank-to-bank transfers used as payments.  So basically, you and the merchant both have a bank account, and you issue a payment from your account to theirs.  You and the merchant never actually touch the Bitcoin ledger (and the bank(s) may not have to either), but it's just a balance transfer on paper.  It wouldn't be much different than the current credit card/debit card system.

We moved to using credit/debit cards for convenience.  Given how inconvenient it is to utilize a Bitcoin client in-person, I can imagine we would move to a similar system with Bitcoins.  It's kind of depressing because Bitcoins were made to move away from all of that stuff, but it's reality.  If Bitcoin usage became widespread, more convenient methods of making payments cropping up would be inevitable.

The convenience of credit cards comes from the inability to use cash online or by mail and the general dangers/issues of purchasing everything w/ cash.

Bitcoin "lack of convenience" is more a technical not fundamental issue.

Imagine something like this:
1) you go shopping and cashier rings up the goods.
2) the little checkout display has a button Bitcoin.
3) you click on it and it displays a QR code.
4) you scan the QR code w/ mobile phone and (based upon an agreed upon format) it has the payment address, amount, and a note for your wallet.
5) you hit send.  3-5 seconds later the cashier gets a payment confirmation, prints the receipt and you walk out of the store.

[SgtSpike]

But see, as long as banks are used, they can manipulate the money supply.  Fractional reserve, and all of that.  If people accept a bank balance as good as coins themselves, then banks can lend out more BTC than they actually have, effectively inflating the BTC supply.  And just as easily, they can withdraw their lending to shore up their reserves and deflate the BTC supply.

I think Bitcoin banks could very easily and credibly prove they have the BTC they say they have via the digital signature function.

The only slightly more difficult part would be them credibly proving that the amount they have on deposit is accurate.  Not impossible though.  I propose a simple straightforward way.

One way they could do this is to create an anonymous list of all their depositors and put them in some sequence.

Example, MtGox could say and sign, "we have 1000000 BTC on deposit", and create an ordered list of balances.  Each entry on the list would have a start and end range, for example, the first entry for example 25 BTC would "start" at 0 and "end" at 25, and the second entry for example 100 BTC would "start" at 25 and "end" at 125.  Each entry would have a public key published with it, and the entire list would be signed by MtGox.  The last item on the list would "end" at 1000000 BTC.

Then MtGox would give each depositor a signed statement that "your deposit, acct xyxy is on our list at <start> and <end>", along with the matching private key for that entry.

Two depositors who could produce messages both claiming the same entry as their deposit could publicly prove that the bank misrepresented its liabilities.

The private and public key scheme included with each line would simply allow the individual depositors to collaborate via some independent registry, so they could anonymously claim ownership of their funds in an effort to look for double allocations without sharing their account numbers or identities at large.  (e.g. they would each sign messages saying "These funds are mine, contact me <how> if you think they're yours", where <how> is generally going to be "publish your contact details encrypted with <public key> and proof that the bank signed a message saying these are yours, and I'll find you.")  While not every depositor would participate, it would only take one collision to discredit the bank - which would be a strong incentive to keep the bank honest.

Makes sense, though it is a rather complex solution to the problem, and I'm not sure many people would understand.  I wonder how many people would actually participate in verifying such a list, or how many companies could be forced into releasing one?  I mean, look at how many people use MtGox already, despite not knowing whether they actually have enough Bitcoins to cover all of their balances?

It might be one of those domino-effect things, where the first company to start regularly releasing such a list might be favored enough by consumers to force other companies to follow suit.  But, I have a feeling not enough people would be concerned about the fractional reserve problem to spend any of their own time on ensuring companies are kept accountable.

Multisig transactions should solve all the banking problems. Your deposits are with them, but they cannot move them unless you provide the other signature. Double spends can be prevented because the bank will not sign a double spend. The "bank" is basically reduced to a trusted third party that prevents double spends, allowing merchants to trust 0-conf transactions.

Seems reasonable.  I have a feeling that banks wouldn't like that much, and many wouldn't use multisigs.  Smart people would, of course, use this method, but we all know how smart the general populace can be...

Some people might not even want to deal with multisigs either.  They like just swiping a card and entering a pin, and unless/until Bitcoin is just as easy (i.e., doesn't require a mobile phone to provide your piece of the multisig), I think many people would be fine with just letting the bank handle the transactions entirely.

But, hopefully you're right, and multisigs would keep fractional banking to a minimum.

Yeah, exactly waspoza.  A merchant would be much more likely to trust a transaction coming directly from a bank than from a consumer standing there with a handheld Bitcoin client in their hand.

The next step, of course, would be bank-to-bank transfers used as payments.  So basically, you and the merchant both have a bank account, and you issue a payment from your account to theirs.  You and the merchant never actually touch the Bitcoin ledger (and the bank(s) may not have to either), but it's just a balance transfer on paper.  It wouldn't be much different than the current credit card/debit card system.

We moved to using credit/debit cards for convenience.  Given how inconvenient it is to utilize a Bitcoin client in-person, I can imagine we would move to a similar system with Bitcoins.  It's kind of depressing because Bitcoins were made to move away from all of that stuff, but it's reality.  If Bitcoin usage became widespread, more convenient methods of making payments cropping up would be inevitable.

The convenience of credit cards comes from the inability to use cash online or by mail and the general dangers/issues of purchasing everything w/ cash.

Bitcoin "lack of convenience" is more a technical not fundamental issue.

Imagine something like this:
1) you go shopping and cashier rings up the goods.
2) the little checkout display has a button Bitcoin.
3) you click on it and it displays a QR code.
4) you scan the QR code w/ mobile phone and (based upon an agreed upon format) it has the payment address, amount, and a note for your wallet.
5) you hit send.  3-5 seconds later the cashier gets a payment confirmation, prints the receipt and you walk out of the store.

But what if you have a custom client on your mobile phone?  Your custom client could do mischevious things such as:
- Create another transaction right after the first one from the same address to one of your own addresses.  It's a crap shoot as to which transaction gets into the block chain at that point.
- Ensure that it is a transaction that would require a fee.  Then, create the transaction without a fee.  It will never go through (or at least, take a long time to go through), but the cashier will see 0-conf.  In the meantime, you can send those coins to a different address to ensure that transaction never happens.
- Etc, etc, etc.  More loopholes will be found, I am certain.

It's no worse than the type of fraud that people can commit with written checks, but you can see how most stores view those these days...

[casascius]

We all hopefully realize that the block chain infrastructure is non-scalable to the level of mass retail adoption and that this discussion is academic at best. By the time retail adoption of bitcoin is commonplace, space on the blockchain will have a price, retail purchases will never hit the blockchain, and you will be dealing with a bank that is simply backed in bitcoins and who settles your purchase off the block chain, who will solve this problem the same way Visa debit works.

Watch Dan Kaminsky's presentation. He agrees.

[SgtSpike]

We all hopefully realize that the block chain infrastructure is non-scalable to the level of mass retail adoption and that this discussion is academic at best. By the time retail adoption of bitcoin is commonplace, space on the blockchain will have a price, retail purchases will never hit the blockchain, and you will be dealing with a bank that is simply backed in bitcoins and who settles your purchase off the block chain, who will solve this problem the same way Visa debit works.

Yeah, exactly, that's what I was initially writing about.

[tvbcof]

We all hopefully realize that the block chain infrastructure is non-scalable to the level of mass retail adoption and that this discussion is academic at best. By the time retail adoption of bitcoin is commonplace, space on the blockchain will have a price, retail purchases will never hit the blockchain, and you will be dealing with a bank that is simply backed in bitcoins and who settles your purchase off the block chain, who will solve this problem the same way Visa debit works.

The most important thing to me is that a system remains firmly within the grasp of the all classes of user-base and can be operated successfully in the face of a dedicated attack by parties who control global network infrastructure.

A solution which I would be OK with would be to pay whatever fee is necessary to keep the block-chain spam-free enough to achieve the objective that I would like to see.

A better solution which I have elaborated on earlier is to simply break the problem into multiple block-chains thus distributing the load and providing better tuned currency solutions in the process (while paying particular attention to protect the 'satoshi block-chain'.)

Watch Dan Kaminsky's presentation. He agrees.

I'll dig it up this weekend.  I've at least read some of his stuff, possibly a slide-show from the presentations you mention, but don't know that I've watched such a thing.

But from your description and my other observations within the community, it strikes me that there is some room for more creative thinking.  I get the strong sense that exceptional computer resources considered a given for problem solving (very common in this space), and also that what I consider a 'problem' is not considered as such by very many people.

[evoorhees]

We all hopefully realize that the block chain infrastructure is non-scalable to the level of mass retail adoption and that this discussion is academic at best. By the time retail adoption of bitcoin is commonplace, space on the blockchain will have a price, retail purchases will never hit the blockchain, and you will be dealing with a bank that is simply backed in bitcoins and who settles your purchase off the block chain, who will solve this problem the same way Visa debit works.

Watch Dan Kaminsky's presentation. He agrees.

+1

In the future, most transactions won't "hit the blockchain" at all. Already today, many don't. Anyone sending money to someone else of the same ewallet may not ever hit the blockchain. Same with transactions occurring within an exchange. Handing a persona  Casascius coin doesn't hit the blockchain either. (Sidenote: this is also why Bitcoin can be untraceable - the blockchain doesn't record every transaction, contrary to popular belief)

The blockchain will always be the ultimate arbiter. When one needs -absolute- confidence in a transaction, one would rely on the blockchain. However, 99% of real-world transactions don't require absolute confidence, they just require -reasonable- confidence, and within that realm we'll see lots of interesting market-based systems to quickly process payments.

This phenomenon doesn't detract from the strengths of Bitcoin. Rather, it's simply the evolving scaffolding which surrounds it.

[evoorhees]

But see, as long as banks are used, they can manipulate the money supply.

How?  Anything a bank/ewallet can do with bitcoin, so can I. We are all, already, banks.

As a bank, I could extend credit to my friends right now, this instant. It could be in the form of an IOU, redeemable in Bitcoin. IOU's for Bitcoin can always be made, copied, and passed around the economy. If there is trust among the market players, then this effectively increases the money supply - but it's not necessarily harmful.

"Fractional reserve banking" is not the enemy. The enemy is fractional reserve banking that is subsidized by the government - for without the subsidization market forces are able to work and fractional banks that act imprudently will be destroyed.

[casascius]

Makes sense, though it is a rather complex solution to the problem, and I'm not sure many people would understand.  I wonder how many people would actually participate in verifying such a list, or how many companies could be forced into releasing one?  I mean, look at how many people use MtGox already, despite not knowing whether they actually have enough Bitcoins to cover all of their balances?

It might be one of those domino-effect things, where the first company to start regularly releasing such a list might be favored enough by consumers to force other companies to follow suit.  But, I have a feeling not enough people would be concerned about the fractional reserve problem to spend any of their own time on ensuring companies are kept accountable.

Not many people would need to participate for it to be effective.  Lots of people already wonder if MtGox has reserves to cover their deposits.  Imagine TradeHill started doing this today.  That would be HUGE news, a major leap in trustability.

Sure, the bank could pull a fast one on some people assuming they won't all check.  But the more a bank overextends itself, the greater the likelihood they are going to get caught with this scheme.  The signature scheme would guarantee that even a single detected instance of double-allocation of funds would be provable.  It would be enough to impeach the bank as dishonest, even if it were for a trivial amount.

[bitcoinbetas]

$10 by this time next week    wanna bet?

[SgtSpike]

But see, as long as banks are used, they can manipulate the money supply.

How?  Anything a bank/ewallet can do with bitcoin, so can I. We are all, already, banks.

As a bank, I could extend credit to my friends right now, this instant. It could be in the form of an IOU, redeemable in Bitcoin. IOU's for Bitcoin can always be made, copied, and passed around the economy. If there is trust among the market players, then this effectively increases the money supply - but it's not necessarily harmful.

"Fractional reserve banking" is not the enemy. The enemy is fractional reserve banking that is subsidized by the government - for without the subsidization market forces are able to work and fractional banks that act imprudently will be destroyed.

I think the level of lending achievable on a personal basis vs the level of lending achievable by a bank are two completely different things though.  A bank could probably fairly easily leverage itself to lend 5x or 10x as much money as it actually has in deposits.  I can't see very many people doing the same thing on a "friend level" basis.

Regardless, I got a bit sidetracked.  I do agree that fractional reserve banking itself isn't the enemy.

Makes sense, though it is a rather complex solution to the problem, and I'm not sure many people would understand.  I wonder how many people would actually participate in verifying such a list, or how many companies could be forced into releasing one?  I mean, look at how many people use MtGox already, despite not knowing whether they actually have enough Bitcoins to cover all of their balances?

It might be one of those domino-effect things, where the first company to start regularly releasing such a list might be favored enough by consumers to force other companies to follow suit.  But, I have a feeling not enough people would be concerned about the fractional reserve problem to spend any of their own time on ensuring companies are kept accountable.

Not many people would need to participate for it to be effective.  Lots of people already wonder if MtGox has reserves to cover their deposits.  Imagine TradeHill started doing this today.  That would be HUGE news, a major leap in trustability.

Sure, the bank could pull a fast one on some people assuming they won't all check.  But the more a bank overextends itself, the greater the likelihood they are going to get caught with this scheme.  The signature scheme would guarantee that even a single detected instance of double-allocation of funds would be provable.  It would be enough to impeach the bank as dishonest, even if it were for a trivial amount.

Agreed.

[tvbcof]

I think the level of lending achievable on a personal basis vs the level of lending achievable by a bank are two completely different things though.  A bank could probably fairly easily leverage itself to lend 5x or 10x as much money as it actually has in deposits.  I can't see very many people doing the same thing on a "friend level" basis.

Regardless, I got a bit sidetracked.  I do agree that fractional reserve banking itself isn't the enemy.

There is a good reason why fractional reserve banking has been ubiquitous for eons.  It has the potential to be very sound given good management, and provides the flexibility to dampen features inherent in an economy.

The key is 'good management'.  As long as there is money to be made by neglecting 'good management', it is unlikely to occur on an ongoing basis.

I believe that management of a monetary system should be devoid of a profit motive and that all of the benefits of having one and using it should accrue to the citizens who use it.  In short, a monetary system should be a public utility.

What we in the USA have with private ownership of the monetary system with state enforced 'tender' laws is the worst of both worlds, and is fascism by my definition.  The construct always had a built in life expectancy due to it's reliance on inflation (which the owners can scrape), and I suspect that we are likely near the it's end-of-life.  The replacement will be as much like what we have now as possible, or even better from the standpoint of those who have been milking the current system.

The success of Bitcoin is important to me in large part because this could provide tangible evidence of the possibility for alternative monetary solutions when they are needed.  If Bitcoin (or something like it) could be chosen against the will of 'the 1%' (for lack of a better description) that would be a great thing, and this is why it is some important to me that the network itself retains it's ability to operate against the types of concerted attacks which seem improbably to most people.

---

Oh ya, I would happily patronize a 'fractional reserve Bitcoin bank' if there were a reason to do so.  My criteria would be that 1) I get a fair cut of the proceeds, 2) I hold the power to get out imediately and autonomously if I choose, and 3) that I can verify their books to my satisfaction.

[FreeMoney]

Oh ya, I would happily patronize a 'fractional reserve Bitcoin bank' if there were a reason to do so.  My criteria would be that 1) I get a fair cut of the proceeds, 2) I hold the power to get out imediately and autonomously if I choose, and 3) that I can verify their books to my satisfaction.

By #2 do you mean that people can get out in order of asking and you'll only be left holding the bag if you are late? Because not everyone can get out, that's the downside.

[tvbcof]

Oh ya, I would happily patronize a 'fractional reserve Bitcoin bank' if there were a reason to do so.  My criteria would be that 1) I get a fair cut of the proceeds, 2) I hold the power to get out imediately and autonomously if I choose, and 3) that I can verify their books to my satisfaction.

By #2 do you mean that people can get out in order of asking and you'll only be left holding the bag if you are late? Because not everyone can get out, that's the downside.

No reward (#1) is without risk (#you).  I would want the tools to protect myself (#3 & #2) before I would risk my assets.  I'm not planning to be MF-Global'd/MyBitcoin'd with Bitcoin or any other asset I hold.

And I'm very unlikely to loan my BTC to a bank/exchange/whatever for a paltry sub-inflation % as I am forced to with the USD I hold (and mostly dump as quickly as I get them.)