Does bitcoin have an expiry date? What do you think?

[manny_el_portu]

From what I understand about bitcoin (which, granted, isn't all that much), it is the mining process which actually makes the system secure by confirming transactions and stopping double spending, right?
If that truly is the case, then it seems to me that bitcoin was created with a natural expiry date.

What I mean is, once the last block has been found and the last coins have been mined, there will be no more incentive to keep mining and "giving" you processing power to make the system secure. This means that everyone who has been mining for a profit will stop doing so. Sure, there may be a few people who continue mining for nothing just for the pleasure of keeping the system running, but most people don't do anything for free, so the hashrate of the network will drop precipitously.

If the hashrate drops low enough (which it almost certainly will) all transactions will take forever to confirm which will make bitcoin impossible to use as an effective currency (who wants to use a currency that requires you to wait 6 months between the moment you make the transaction and the moment it is actually confirmed?).

This unusability, in turn, should make the price of bitcoins drop enormously instead of going up, as making it impossible to use bitcoins for actual transactions should push all hoarders to try to dump their now useless coins, creating a great increase in supply accompanied by either no change or a decrease in demand.

If all this happens (and it makes sense to me that it will) bitcoin has essentially killed itself by having a finite number of minable coins. What do you think?

[ProfMac]

From what I understand about bitcoin (which, granted, isn't all that much), it is the mining process which actually makes the system secure by confirming transactions and stopping double spending, right?
If that truly is the case, then it seems to me that bitcoin was created with a natural expiry date.

What I mean is, once the last block has been found and the last coins have been mined, there will be no more incentive to keep mining and "giving" you processing power to make the system secure. This means that everyone who has been mining for a profit will stop doing so. Sure, there may be a few people who continue mining for nothing just for the pleasure of keeping the system running, but most people don't do anything for free, so the hashrate of the network will drop precipitously.

If the hashrate drops low enough (which it almost certainly will) all transactions will take forever to confirm which will make bitcoin impossible to use as an effective currency (who wants to use a currency that requires you to wait 6 months between the moment you make the transaction and the moment it is actually confirmed?).

This unusability, in turn, should make the price of bitcoins drop enormously instead of going up, as making it impossible to use bitcoins for actual transactions should push all hoarders to try to dump their now useless coins, creating a great increase in supply accompanied by either no change or a decrease in demand.

If all this happens (and it makes sense to me that it will) bitcoin has essentially killed itself by having a finite number of minable coins. What do you think?

The miner's fee is a source of income.  It will gradually become more important as the number of mined coins becomes smaller.

[manny_el_portu]

Miner's fee? I had no idea that existed (or does it not exist yet?) Granted, that would solve the problem, but it would have to be a high enough fee to both pay for the cost of mining and still give a profit to the miner. How much is the miner's fee?

[Deafboy]

https://blockchain.info/charts/transaction-fees

[DannyHamilton]

Miner's fee? I had no idea that existed (or does it not exist yet?) Granted, that would solve the problem, but it would have to be a high enough fee to both pay for the cost of mining and still give a profit to the miner. How much is the miner's fee?

Miner's fees (frequently called "transaction fees") have always existed.  They are generally voluntary, but in some cases some wallets enforce them.  As a general rule right now, the typical transaction fee is 0.0001 BTC per kilobyte of transaction size.  Users can voluntarily include a larger transaction fee to provide an incentive to miners to confirm their transaction sooner.

The block reward paid to miners is the sum of the block subsidy and all the transaction fees from all the transactions that the miner includes in the block.  At the moment the subsidy is 25 BTC, so the total of the transaction fees is a very small percentage of the total block reward.  However, every 4 years the block subsidy is cut in half.  Meanwhile as bitcoin becomes more popular and more people use it, the value of the transaction fees increases.  Eventually the transaction fees in the block will exceed the subsidy.  While the process will remain exactly the same, perhaps we'll stop calling it "mining" and start calling it "transaction processing" at that time?  Regardless, the subsidy will continue to shrink.  a bit over 130 years from now, the subsidy will finally shrink to 0 bitcoins, and by then the entire block reward will be from transaction fees.

[quellerdrive]

Miner's fees (frequently called "transaction fees") have always existed.  They are generally voluntary, but in some cases some wallets enforce them.  As a general rule right now, the typical transaction fee is 0.0001 BTC per kilobyte of transaction size.  Users can voluntarily include a larger transaction fee to provide an incentive to miners to confirm their transaction sooner.

The block reward paid to miners is the sum of the block subsidy and all the transaction fees from all the transactions that the miner includes in the block.  At the moment the subsidy is 25 BTC, so the total of the transaction fees is a very small percentage of the total block reward.  However, every 4 years the block subsidy is cut in half.  Meanwhile as bitcoin becomes more popular and more people use it, the value of the transaction fees increases.  Eventually the transaction fees in the block will exceed the subsidy.  While the process will remain exactly the same, perhaps we'll stop calling it "mining" and start calling it "transaction processing" at that time?  Regardless, the subsidy will continue to shrink.  a bit over 130 years from now, the subsidy will finally shrink to 0 bitcoins, and by then the entire block reward will be from transaction fees.

This is the perfect answer to the question I had in mind earlier as I had wondered about what would happen once all 21 million BTCs were mined and what role "miners" would still play in the BTC network. So actually, miners essentially become a pool of data processors for BTC transactions. What does it mean when a block is rejected and how does that come about? Through what mechanism in the BTC network does a higher transaction fee added to a transaction would prompt faster processing / confirmation to a transaction?

[ExoCoin]

BTC in theory should last forever but if it reaches the 21 million and there are not many transactions what is going to happen? Less miner = Less secure and if it get's attacked it could lose a lot of users and therefor essentially killing it.

[DannyHamilton]

In short, how would you summarize what is required for and would constitute a (transaction) confirmation

To "confirm" a transaction (to give it it's first confirmation), a miner verifies that the transaction conforms to the rules of the protocol, then they package the transaction together with other verified transactions into a block generating a block header that is unique to that list of transactions, then they search for a nonce that results in the hash of the header being lower than the current difficulty, then they broadcast the block for peers to validate and add to their blockchain.  When a transaction is first included in a block in the blockchain, that transaction has its first "confirmation".

and does it require a certain number of miners to "confirm" it?

Only 1 miner will confirm any particular transaction.

I understand that the number of confirmation of a transaction is merely the number of new blocks that have been generated after the transaction is "posted"(?) to the blockchain, so is it because the more new blocks posted since a given transaction that it quickly becomes extremely difficult for that given transaction to be altered in the blockchain and hence the transaction is nailed down securely for good into the blockchain?

Correct.  The hash of the transaction is included in the block header as a merkle root (in other words, change any transaction in a block, and you change the header of the block).  In doing so, the nonce that was published with the block would no longer result in the published hash of the block header.  To convince peers to replace the block in the blockchain, an attacker would have to mine a new block that has a low enough hash value.  Each block has in its header, the hash of the previous block, so to change a block that has a new block added after it, an attacker would have to re-mine both blocks to find valid new nonces for each.  The further back an attacker tries to modify a block, the more newer blocks the attacker has to re-mine to convince peers to accept it.  Meanwhile, the rest of the honest network is mining new blocks, so the attacker has to re-mine the old blocks faster than the rest of the network is creating new ones or they'll never catch up (and peers will only replace their current blockchain if the replacement is "longer" than the current one).  Of course if an attacker could provide more hashing power than the entire combined honest network, then the attacker would be guaranteed that they would catch up eventally.  That's typically referred to as a "51% attack".

Also, what does it mean when a block is rejected and how does that come about?

No sure what exactly you are referring to.  If an attacker were to try to broadcast an invalid block, then all peers would simply refuse to relay it or add it to their own copy of the blockchain.  If two miners find solutions for two different blocks (that each have the same block hash listed as the "previous block"), then peers will accept the first of those two blocks that are relayed to them.  When they receive the second block that has the same "parent", they will reject it because it creates a chain that is the same length, and acceptance of replacements to current blocks requires a "longer" chain.  This can result in a temporarily "split" network, where some peers have one block added to their blockchain, and others have the other block.  Eventually a miner working on one side of the split or the other will find a "next" block.  When this new block is broadcast, it will result in a chain that is now the "longer" of the two.  Therefore, peers that had been on the other side of the "split" will reject (orphan) the block they currently have and replace it with the two new blocks of the now longer chain.

Through what mechanism in the BTC network does a higher transaction fee added to a transaction would prompt faster processing / confirmation to a transaction?

There is nothing in the protocol that requires a miner to include anyone's transactions in their block.  A miner includes transactions for two reasons.

The first reason is that if no transactions ever get included then bitcoin becomes useless and the mined bitcoins are therefore useless.  The miners add value to all bitcoins by making them useful.  The miners therefore add value to their own bitcoins by making all bitcoins useful.  Now that miners have an incentive to add some transactions to the blocks that they create, the question is which ones to add?

The second reason is that they get paid to do so with transaction fees. Transaction fees provide an incentive for miners to add a particular transaction.  In the beginning there weren't many transaction fees.  Miners had very few transactions to choose from, and plenty of space available for those transactions within the one megabyte limit that the protocol imposes on each block.  As such, miners were happy to include all available transactions in the block they were working on just to make bitcoin useful.  As bitcoin becomes more popular, there can be more unconfirmed transactions than available space within the one megabyte limit.  At that point, a miner has to decide which transactions to include in the block they are working on.  By adding a transaction fee, you provide a financial incentive to the miners to include your transaction in their block instead of some other transaction that paid a smaller fee.  All the transactions that don't make it into the next solved block have to wait unconfirmed until some other block where there is space for them.  Unconfirmed transactions are easier for an attacker to modify since they aren't protected by the hashing work that miners do, so there is an incentive to users to get their transactions confirmed as quickly as possible.

[DannyHamilton]

BTC in theory should last forever but if it reaches the 21 million and there are not many transactions what is going to happen? Less miner = Less secure and if it get's attacked it could lose a lot of users and therefor essentially killing it.

Either bitcoin will gain popularity, in which case there will be plenty of transactions to provide an incentive to miners via transaction fees, or bitcoin will die long before it gets to 20,999,999.9769 bitcoins. Why would miners pay for all the electricity necessary to mine during the 4 years before the last bitcoin is mined when the subsidy is 0.00000001 BTC, unless there were enough transaction fees to make up the difference in revenue.  Why would they mine during the 4 years before that when the subsidy is only 0.00000002 BTC without sufficient transaction fees, or the 4 years before that when the subsidy is only 0.00000004 BTC?

As you can see, either the popularity (and therefore number of transactions) of bitcoin increases over time generating enough revenue from transaction fees, or it doesn't and bitcoin dies long before the last bitcoin is mined.

[quellerdrive]

Either bitcoin will gain popularity, in which case there will be plenty of transactions to provide an incentive to miners via transaction fees, or bitcoin will die long before it gets to 20,999,999.9769 bitcoins. Why would miners pay for all the electricity necessary to mine during the 4 years before the last bitcoin is mined when the subsidy is 0.00000001 BTC, unless there were enough transaction fees to make up the difference in revenue.  Why would they mine during the 4 years before that when the subsidy is only 0.00000002 BTC without sufficient transaction fees, or the 4 years before that when the subsidy is only 0.00000004 BTC?

As you can see, either the popularity (and therefore number of transactions) of bitcoin increases over time generating enough revenue from transaction fees, or it doesn't and bitcoin dies long before the last bitcoin is mined.

Even as of this month August 2013, the current mining difficulty level has reached 33 at the moment. Most available mining hardware except the most powerful ones out there will not breakeven or produce minimal profit after expenses (mining hardware costs / electricity) based on mining calculators out there. It seems mining (for subsidies) is pretty much almost over within the next couple of months, unless you got a headstart with either GPU/FPGA rigs earlier, or an ASIC machine in early 2013. Those few that did would have been handsomely rewarded this year. The increase in difficulty level is also accelerating, so it seems miner's future revenue will rely heavily on transaction fees very soon but it's unclear at this point if the BTC currency will be adopt for use fast enough for TX fees to make up in lost mining subsidy revenue. In this transaction fee graph posted earlier in this thread:

https://blockchain.info/charts/transaction-fees

would the transaction fee value at any given point on the graph be that of the TX fee on any given day for all miners in the entire BTC network? It would be nice if there was a calculator on line somewhere that could give us an idea of TX fee as a function of mining hardware power which could be inputted by a user which had mining hardware.

[DannyHamilton]

Even as of this month August 2013, the current mining difficulty level has reached 33 at the moment. Most available mining hardware except the most powerful ones out there will not breakeven or produce minimal profit after expenses (mining hardware costs / electricity) based on mining calculators out there. It seems mining (for subsidies) is pretty much almost over within the next couple of months, unless you got a headstart with either GPU/FPGA rigs earlier, or an ASIC machine in early 2013. Those few that did would have been handsomely rewarded this year. The increase in difficulty level is also accelerating, so it seems miner's future revenue will rely heavily on transaction fees very soon but it's unclear at this point if the BTC currency will be adopt for use fast enough for TX fees to make up in lost mining subsidy revenue.

If mining difficulty gets too high to be profitable, then some miners will shut off their mining equipment. If enough miners shut their equipment off, then difficulty will drop. Eventually it will drop far enough for mining to be profitable, and miners will start turning equipment back on. An equilibrium will be reached where difficulty allows only the most efficient miners to remain profitable.

This is the intentional design to maximize the security:cost ratio.

[manny_el_portu]

Considering the hashrate of ASICs being produced right now I find it difficult to believe that enough GPU and low end ASICs will be shut down in order to make an actual difference. I believe a better way to provide an incentive to mining would be to charge a small percentage of each transaction (say 1% or 0,5%) that would be credited to the miner/miners that confirmed the transaction. This could really make the network bloom and would provide an incentive for even the lowest of hasrahte miners to stay online. Someone should suggest that to Satoshi Nakamoto...

[DannyHamilton]

I believe a better way to provide an incentive to mining would be to charge a small percentage of each transaction (say 1% or 0,5%) that would be credited to the miner/miners that confirmed the transaction. This could really make the network bloom and would provide an incentive for even the lowest of hasrahte miners to stay online. Someone should suggest that to Satoshi Nakamoto...

I agree with you, but I think we should make the percentage voluntary.  Those users who want their transactions in the very next block can voluntarily pay a higher percentage.  Those who are willing to wait until later blocks can voluntarily pay a lower percentage.  I'll grab my time machine, travel back to 2008 and suggest it to Mr. Nakamoto.  I'll be right back...

Ok, I'm back, did it work?  Did he implement a system that allows users to voluntarily pay a percentage of their transaction to the miner/miners that confirm the transaction?

[ProfMac]

I believe a better way to provide an incentive to mining would be to charge a small percentage of each transaction (say 1% or 0,5%) that would be credited to the miner/miners that confirmed the transaction. This could really make the network bloom and would provide an incentive for even the lowest of hasrahte miners to stay online. Someone should suggest that to Satoshi Nakamoto...

I agree with you, but I think we should make the percentage voluntary.  Those users who want their transactions in the very next block can voluntarily pay a higher percentage.  Those who are willing to wait until later blocks can voluntarily pay a lower percentage.  I'll grab my time machine, travel back to 2008 and suggest it to Mr. Nakamoto.  I'll be right back...

Ok, I'm back, did it work?  Did he implement a system that allows users to voluntarily pay a percentage of their transaction to the miner/miners that confirm the transaction?

Wow!  That was so seamless I don't even remember a time when it wasn't true!

[DannyHamilton]

I believe a better way to provide an incentive to mining would be to charge a small percentage of each transaction (say 1% or 0,5%) that would be credited to the miner/miners that confirmed the transaction. This could really make the network bloom and would provide an incentive for even the lowest of hasrahte miners to stay online. Someone should suggest that to Satoshi Nakamoto...

I agree with you, but I think we should make the percentage voluntary.  Those users who want their transactions in the very next block can voluntarily pay a higher percentage.  Those who are willing to wait until later blocks can voluntarily pay a lower percentage.  I'll grab my time machine, travel back to 2008 and suggest it to Mr. Nakamoto.  I'll be right back...

Ok, I'm back, did it work?  Did he implement a system that allows users to voluntarily pay a percentage of their transaction to the miner/miners that confirm the transaction?

Wow!  That was so seamless I don't even remember a time when it wasn't true!

Time travel is like that.  Only the traveler remembers both realities.

[quellerdrive]

If mining difficulty gets too high to be profitable, then some miners will shut off their mining equipment. If enough miners shut their equipment off, then difficulty will drop. Eventually it will drop far enough for mining to be profitable, and miners will start turning equipment back on. An equilibrium will be reached where difficulty allows only the most efficient miners to remain profitable.

This is the intentional design to maximize the security:cost ratio.

I'm sure what you said is right, and indeed in reviewing past history of difficulty level, I did notice some drop of difficulty some time back (before 2013), but it really didn't drop too much. I did some simulations / calculations with this great calculator below (which by the way, currently is inserting an incorrect difficulty of about 26, and you'll have to manually type in 37 as the correct current difficulty):

http://mining.thegenesisblock.com/

Using one of the most cost effective available ASIC hardware for instance, at 400 GH/s, there is a huge difference (i.e. drop) in mining profitability for each of the months from August to November in 2013, once you punch in all the correct numbers but just varying the starting date of mining. It seems that essentially by November 2013, there is no existing announced (even if unshipped) hardware that will be available to breakeven and/or produce a substantial worthwhile mining subsidy profit. The situation looks very grim on that front. Unless I have plugged in my numbers incorrectly or missed something, any (ASIC) mining starting around early to spring 2014 will require at least a 10 to 15 fold increase in computing power / hashing speed to restore profitiability to any substantially worthwhile level (over the operation of the equipment at least for a stretch of several months). Meanwhile, it seems that there's this mad rush for everyone to get on ASIC equipment (as well as the huge back log of orders from some manufacturers yet to be shipped) and the network will continue to be flooded with them for at least the next while, keeping difficulty level high and mining revenues minimal to non-existent (after deducting operating expenses).

[manny_el_portu]

To all the time travelling replies, I actually thought that the idea could be implemented, I thought fixes could be made to the network in case bugs were found, and I thought my suggestion, if it could be implemented, was actually a good idea.
As for quellerdrive's reply, that is indeed a very serious concern. If the calculations are correct, it seems that no hardware shipping from november onward will even be able to break even. Either the hardware's price drops steeply or we could be looking at the death of bitcoin in the very near future. Either that or the price of bitcoins has to start rising A LOT in order to make it worthwhile for people to keep mining. Even KnC's Jupiter is set to start costing more than it's making as of June 2014

[xan_The_Dragon]

The amount of bitcoins increases by a decelllerating rate to almost 0 at 21 million bitcoins So no

[quellerdrive]

As for quellerdrive's reply, that is indeed a very serious concern. If the calculations are correct, it seems that no hardware shipping from november onward will even be able to break even. Either the hardware's price drops steeply or we could be looking at the death of bitcoin in the very near future. Either that or the price of bitcoins has to start rising A LOT in order to make it worthwhile for people to keep mining. Even KnC's Jupiter is set to start costing more than it's making as of June 2014

Yes, it looks very grim after the next 2 months or so due to the accelerating rate of difficulty increase. What I found is that in that mining calculator (which is very useful because it maps out each month's projected difficulty rate) is that you would have to increase the computing power of the mining hardware well into the multi-terahash range and beyond before any substantial mining revenue can be realized for any mining equipment purchased in and started within early 2014. It seems that even KNC units shipped in October are pretty screwed in terms of revenue, especially those shipping in late October, which would mean that October 2013 is essentially a lost month and factoring in any possible technical delays in the user setting up, November 2013 would not be an inconceivable month for late October shipped KNC units to actually start mining. Well, you can change the starting mining date in the calculator and see the results, which will show a drastically reduced profit compared to a start mining date in September 2013. I saw some mining hardware manufacturers (or was it ASIC chip manufacturers) have month based pricing which means they take into account the massive loss of mining revenue even with just a single passing month. I'm not hopeful that mining equipment could improve in computing power by 10 to 15 fold by early spring 2014, and if not, there's not enough power to even break even for all the trouble involved, unless the value of BTC suddenly doubles and holds at the $200 range or higher - not holding my breath on that one. In reality, it seems the ones that really cashed in big time with ASIC technology are those 900 buyers of Avalon batch one and two, and that hardware at about 60 MH/s is already almost too weak at the present level of difficulty being at 37M.

[Foxpup]

To all the time travelling replies, I actually thought that the idea could be implemented, I thought fixes could be made to the network in case bugs were found, and I thought my suggestion, if it could be implemented, was actually a good idea.

The joke is that this idea is already implemented, and always has been. That's how time travel works. A basic principle of time travel is that once an event in the past is changed, everything in the present will change as result, in the same way that events in the present change the future. Included in these changes is everyone's memories of events: people no longer remember the "original" history of event, because as far as their timeline is concerned, it never happened. Nobody here remembers a time when transaction fees weren't a normal and expected part of the mining incentive. The only record we have of the original history is your forum post, which survived because telecommunications devices behave, well, differently when it comes to time travel, for reasons which are still unclear. That's why Professor Reg could never get his telephone to work.