The fundamental flaw in bitcoin

[Beepbop]

Aren't you? If you're going to do off-chain transactions, you can't actually give away coins by giving people the private key to a wallet. You'll be trading tokens that say "this token is worth 1 BTC" and "this token is worth 0.5 BTC". Even if you can do that securely, at some point you have to trust somebody to hold that coin and cash it in for you. So it's much like one of those countries who have a currency with a fixed parity with a base currency. Or you could call it scrip, gift cards or "barter Bitcoins". In any case you'd have a central (or de-centralized) issuer of these tokens, and you trust them not to empty the tokens of bitcoins (if the token refers to a particular wallet) or you trust them to have reserves of Bitcoin in cold storage to pay you if you want to "cash in" your token.

Edit: Alternatively, you might have some form of DRM/TPM system that makes it possible to trade active bitcoin wallets in a tamper-proof way, so that you know that nobody looked at the private key yet. In that case, sure you could do offline Bitcoin transfer that way, but that's a more advanced system than what we have now. As I said, it might be more fruitful to use it to transfer actual USD if you have such a good system.

[bitlancr]

The problem with what you're suggesting is that once anyone gets a look at the private key, that key becomes worthless to other people. The only way around it is to have a centralized system (a bank), or a peer to peer system for exchanging private keys securely (something like Bitcoin²?). Both of these would be subject to transaction fees of some sort, to pay for the network.

[edd]

"1) instead of being cheap, transactions are now totally free" - Would the smartcards be free? I didn't have to purchase anything to start using bitcoins (unless you count the BTC itself).

"2) instead of taking 10min-1hr to verify, transactions are now instant" - It's only instant because you're not verifying anything!

"3) hackers personally enjoy coding stuff like this" - I'm sure Satoshi, Gavin, and other developers enjoy what they're doing.

"4) a bitcoin business can use it as a selling point" - Businesses could use the opposite as a selling point as well. So far, no one has joined this thread in support of your idea which tells me that there would be customers not interested in it.

"5) transaction can now be totally anonymous" - They can be anonymous now.

"6) it prevents block chain data growing wildly since less transaction processed on net (someone might implement with this good intention without thinking about its long term consequence)" - Not sure I understand this. How would this help anyone play SatoshiDice?

"7) storing a wallet such that is interlinks with a computer's TPM is more secure then storing it on normally on the harddrive" - Many security conscious bitcoiners already avoid storing their BTC on their harddrive.

"8.) someone might actually do this off-network exchange using a smart cards (there have been various proposals already about bitcoin on smart cards) making it easier to carry the wallets around on your person (a lot easier that carrying around your computer)" - Easier than carrying around my phone?

"9) smartcard storages make it easier to store the wallet in a physically secure place (such as a bank vault)" - Easier than a paper wallet which I can do now?

"10) electronic off-net exhange means people without internet but local connectivity (eg: blue-tooth/NFC) can transaction business" - I admit this would be an advantage in those ever-shrinking locations without internet access. Personally, I avoid those like the plague.

"11) phsyical off-network exchange, like casasius coin, allows people without even a  computer" - My phone is often my computer. It seems to me the percentage of the population without near continuous access to sufficient computing power and the internet is rapidly shrinking.

"12) it could be overall easier for people new to bitcoin to learn to use-- since technology like smartcards  and there useage are quite common (such as bus-passes) and the fact that smart card bitcoin would be very similar is use." - Many bitcoiners are working very hard on the issue of making BTC more attractive to the average person. While your idea of a bitcoin smartcard might come to fruition, I don't see more than a few people willing to sacrifice the security and decentralization of the blockchain for something that may be slightly more convenient under certain circumstances.

[Beepbop]

On that last point search for "Woolong device" on these forums. It was a miserable failure. See also http://www.woolong.com/

[beeblebrox]

Edit: Alternatively, you might have some form of DRM/TPM system that makes it possible to trade active bitcoin wallets in a tamper-proof way, so that you know that nobody looked at the private key yet. In that case, sure you could do offline Bitcoin transfer that way, but that's a more advanced system than what we have now. As I said, it might be more fruitful to use it to transfer actual USD if you have such a good system.

That is exactly what I'm talking about..  No, we aren't quite there yet (I never said we were, I said in the future), but we are getting very close.  The majority of new computers have quite advanced DRM systems-- I believe the new windows actually mandates it.  Very soon the technology will be commonplace and used by default (the media content companies are pushing really hard for it since it will drastically reduce copyright infringement).  Once it is common, then I'm willing to bet that someone will use it to exchange bitcoins.

[beeblebrox]

"1) instead of being cheap, transactions are now totally free" - Would the smartcards be free? I didn't have to purchase anything to start using bitcoins (unless you count the BTC itself).

"2) instead of taking 10min-1hr to verify, transactions are now instant" - It's only instant because you're not verifying anything!

"3) hackers personally enjoy coding stuff like this" - I'm sure Satoshi, Gavin, and other developers enjoy what they're doing.

"4) a bitcoin business can use it as a selling point" - Businesses could use the opposite as a selling point as well. So far, no one has joined this thread in support of your idea which tells me that there would be customers not interested in it.

"5) transaction can now be totally anonymous" - They can be anonymous now.

"6) it prevents block chain data growing wildly since less transaction processed on net (someone might implement with this good intention without thinking about its long term consequence)" - Not sure I understand this. How would this help anyone play SatoshiDice?

"7) storing a wallet such that is interlinks with a computer's TPM is more secure then storing it on normally on the harddrive" - Many security conscious bitcoiners already avoid storing their BTC on their harddrive.

"8.) someone might actually do this off-network exchange using a smart cards (there have been various proposals already about bitcoin on smart cards) making it easier to carry the wallets around on your person (a lot easier that carrying around your computer)" - Easier than carrying around my phone?

"9) smartcard storages make it easier to store the wallet in a physically secure place (such as a bank vault)" - Easier than a paper wallet which I can do now?

"10) electronic off-net exhange means people without internet but local connectivity (eg: blue-tooth/NFC) can transaction business" - I admit this would be an advantage in those ever-shrinking locations without internet access. Personally, I avoid those like the plague.

"11) phsyical off-network exchange, like casasius coin, allows people without even a  computer" - My phone is often my computer. It seems to me the percentage of the population without near continuous access to sufficient computing power and the internet is rapidly shrinking.

"12) it could be overall easier for people new to bitcoin to learn to use-- since technology like smartcards  and there useage are quite common (such as bus-passes) and the fact that smart card bitcoin would be very similar is use." - Many bitcoiners are working very hard on the issue of making BTC more attractive to the average person. While your idea of a bitcoin smartcard might come to fruition, I don't see more than a few people willing to sacrifice the security and decentralization of the blockchain for something that may be slightly more convenient under certain circumstances.

I'll grant you point 9, but nothing else.  (By-the-way: these points weren't though out in depth they were on-the-spot thinking. So, the fact that I got one wrong is not surprising)

[Come-from-Beyond]

Please explain to be why this is not considered the major problem with bitcoin and why it is not discussed on this forum.

Bitcoin has a lot of major problems, but a lot of Bitcointalk users own bitcoins and want to earn a lot of money with its help. The only way to do so is to sell the coins to newbies at the highest price. So they avoid discussions that could scare newbies away or make the price to go down.

Of coz, they are still nice ppl. They tell u "Do not invest anything in Bitcoin you can not afford to lose" for conscience sake and this lets them to sleep well.

[edd]

I'll grant you point 9, but nothing else.  (By-the-way: these points weren't though out in depth they were on-the-spot thinking. So, the fact that I got a bit wrong is not surprising)

Why won't you grant anything but point 9?  How were my rebuttals insufficient?

[Beepbop]

That is exactly what I'm talking about..
(.....)
 Once it is common, then I'm willing to bet that someone will use it to exchange bitcoins.

I see your point more clearly now. Call it "Offline tamper resistant exchange of Bitcoin wallets using trusted computing: A system that guarantees nobody before you has looked at the private key." or something like that.
Unfortunately, I think it either might have to be built into the protocol itself, or we're back to a central issuer again. Since the bitcoin client lets you know the private key when you generate your wallet, you'd need some trusted third party to generate the exchangable wallets for you, and provide you with some system that guarantees that nobody else looked at the private key before you "break the seal" and spend it.

So we're back to trusting a third party (some kind of bank) or rewriting the current protocol. It might be feasible in the future, but I think the people who are working on electronic cash (denominated in actual USD) will get there before Bitcoin does....

[niko]

OP raises an interesting point. Why pay mining fees, when you can perform off-chain transactions?  One way to answer this is: because off-chain transactions inevitably involve some cost and some risk. Paying a transaction fee to the miner may be cheaper, faster, and more secure than performing an off-chain transaction. This "may be" is what will create a competitive market that defines the fees depending on particular payer's needs.

[potato5491]

Hello, beeblebrox, I am also new here.

ie: The transaction fee model will not support the cost of block creation.  Bitcoin will lose hashing power in the future when block reward coin generation drops off faster than the increase in bitcoin price.

Why wouldn't the tx fee model work? Surely this is a market forces thing. Miners won't operate (for extended periods of time) at a loss, so they will only accept tx's with sufficiently high fees attached. Tx fees will therefore be dictated by the efficiency with which miners can operate. Miners (or more likely mining farms) with low overheads will be more competitive and can therefore accept lower fees, stealing market share from their competitors.

[beeblebrox]

That is exactly what I'm talking about..
(.....)
 Once it is common, then I'm willing to bet that someone will use it to exchange bitcoins.

I see your point more clearly now. Call it "Offline tamper resistant exchange of Bitcoin wallets using trusted computing: A system that guarantees nobody before you has looked at the private key." or something like that.
Unfortunately, I think it either might have to be built into the protocol itself, or we're back to a central issuer again. Since the bitcoin client lets you know the private key when you generate your wallet, you'd need some trusted third party to generate the exchangable wallets for you, and provide you with some system that guarantees that nobody else looked at the private key before you "break the seal" and spend it.

So we're back to trusting a third party (some kind of bank) or rewriting the current protocol. It might be feasible in the future, but I think the people who are working on electronic cash (denominated in actual USD) will get there before Bitcoin does....

You don't require a trusted third party.  Remote attestion takes care of that.  With remote attestion, I can be guaranteed that your computer is running a particular piece of software.  That software in this case would be the wallet generating one-- if it is open source then I can review the code to make sure that it nevers gives out the key, but only puts it in the wallet which gets passed around.  If this wallet is tampered with to reveal the key then the DRM software reveals to the next person receiving the wallet that it has been tampered with.

It dosen't require any change to the bitcoin protocol to do this.

[beeblebrox]

Hello, beeblebrox, I am also new here.

ie: The transaction fee model will not support the cost of block creation.  Bitcoin will lose hashing power in the future when block reward coin generation drops off faster than the increase in bitcoin price.

Why wouldn't the tx fee model work? Surely this is a market forces thing. Miners won't operate (for extended periods of time) at a loss, so they will only accept tx's with sufficiently high fees attached. Tx fees will therefore be dictated by the efficiency with which miners can operate. Miners (or more likely mining farms) with low overheads will be more competitive and can therefore accept lower fees, stealing market share from their competitors.

It doesn't work because off-chain transaction happen independant of the miners.  There are physical coins call casascius coins that have a bitcoin stored in them. (you can buy them off the internet, search for casascius coin)-- you can phyiscally give these coins to someone who can then give to someone else who can then give them to someone else,   etc......  none of the transactions are being recorded in the block chain and hence there is no fee involved.

[edd]

Hello, beeblebrox, I am also new here.

ie: The transaction fee model will not support the cost of block creation.  Bitcoin will lose hashing power in the future when block reward coin generation drops off faster than the increase in bitcoin price.

Why wouldn't the tx fee model work? Surely this is a market forces thing. Miners won't operate (for extended periods of time) at a loss, so they will only accept tx's with sufficiently high fees attached. Tx fees will therefore be dictated by the efficiency with which miners can operate. Miners (or more likely mining farms) with low overheads will be more competitive and can therefore accept lower fees, stealing market share from their competitors.

It doesn't work because off-chain transaction happen independant of the miners.  There are physical coins call casascius coins that have a bitcoin stored in them. (you can buy them off the internet, search for casascius coin)-- you can phyiscally give these coins to someone who can then give to someone else who can then give them to someone else,   etc......  none of the transactions are being recorded in the block chain and hence there is no fee involved.

That doesn't answer the question. Casascius coins exist and transaction fees are still being paid. There are advantages and disadvantages to using off-chain transactions. You have yet to demonstrate how the advantages outweigh the disadvantages to anyone's satisfaction.

[beeblebrox]

Hello, beeblebrox, I am also new here.

ie: The transaction fee model will not support the cost of block creation.  Bitcoin will lose hashing power in the future when block reward coin generation drops off faster than the increase in bitcoin price.

Why wouldn't the tx fee model work? Surely this is a market forces thing. Miners won't operate (for extended periods of time) at a loss, so they will only accept tx's with sufficiently high fees attached. Tx fees will therefore be dictated by the efficiency with which miners can operate. Miners (or more likely mining farms) with low overheads will be more competitive and can therefore accept lower fees, stealing market share from their competitors.

It doesn't work because off-chain transaction happen independant of the miners.  There are physical coins call casascius coins that have a bitcoin stored in them. (you can buy them off the internet, search for casascius coin)-- you can phyiscally give these coins to someone who can then give to someone else who can then give them to someone else,   etc......  none of the transactions are being recorded in the block chain and hence there is no fee involved.

That doesn't answer the question. Casascius coins exist and transaction fees are still being paid. There are advantages and disadvantages to using off-chain transactions. You have yet to demonstrate how the advantages outweigh the disadvantages to anyone's satisfaction.

Yes, a fee gets paid.  However, it ONLY gets paid ONCE (at the time when casasuis loads the coin) all other subsequent transactions that involve the physical exchange of coins are *completely* free!

[edd]

Yes, a fee gets paid.  However, it ONLY gets paid ONCE (at the time when casasuis loads the coin) all other subsequent transactions that involve the physical exchange of coins are *completely* free!

No, you misunderstood. I'm saying, Casacius Coins exist yet they haven't supplanted the current system.

[beeblebrox]

OP raises an interesting point. Why pay mining fees, when you can perform off-chain transactions?  One way to answer this is: because off-chain transactions inevitably involve some cost and some risk. Paying a transaction fee to the miner may be cheaper, faster, and more secure than performing an off-chain transaction. This "may be" is what will create a competitive market that defines the fees depending on particular payer's needs.

On chain transactions won't be cheaper.  It definitely won't be faster (some transaction currently take upto an hour to complete)- electronic off-chain transactions would be instant. It *just* might be more secure,  however, the vast number of transactions involve small amounts of money (ie: <$20) so who really cares about it being super,super,duper secure compared to super,duper secure when the amount it small?  

Let's be realistic here.  Nobody, is going to wait around for 10mins+ to confirm a transaction for a cup of coffee and a donut that is 99.999999% secure when you have the option for an instant exchange that is 99.999%.

[Aahzman]

I'll nuke your little account out of spite.

Where can I donate to make this feature a reality? Make it a max donation of BTC0.01 per donator, but when the donation for a particular offender hits BTC1.0, KABOOOOOOOOM.  Could be a good way to generate some income for the board. :-)

[potato5491]

It doesn't work because off-chain transaction happen independant of the miners.  There are physical coins call casascius coins that have a bitcoin stored in them. (you can buy them off the internet, search for casascius coin)-- you can phyiscally give these coins to someone who can then give to someone else who can then give them to someone else,   etc......  none of the transactions are being recorded in the block chain and hence there is no fee involved.

I do accept the existence of off-chain tx but IMO off-chain tx will always represent a small minority of all tx, for risk reasons. If I were to become the lucky recipient of 50,000BTC one day, I would certainly want that to be recorded within the block chain. If I were given a physical coin worth 50,000BTC what assurance would I have that the other party doesn't have an identical coin in his pocket? He can then potentially claim the funds within the network before me.

[beeblebrox]

Yes, a fee gets paid.  However, it ONLY gets paid ONCE (at the time when casasuis loads the coin) all other subsequent transactions that involve the physical exchange of coins are *completely* free!

No, you misunderstood. I'm saying, Casacius Coins exist yet they haven't supplanted the current system.

They haven't supplanted the current system becuase they are physical.  A physical coin at the current adoption rates of bitcoin is pretty useless- you can't use it anywhere because nobody physically located near you accepts them (bitcoin is almost exclusively used the moment because it can be used over the internet to transact with remote entities).  Also they require extra expense (you pay a premium to get them).

What I'm talking about it in 2,3,4+ years time when some bright spark invents an electronic version of casascius coin.