Looks like Mt. Gox is dead...

[Mark R]

Thanks for taking the time to answer my queations.

How quickly does the Bitcoin network register the change of ownership?

If I have 5 copies of my Bitcoin and click on 5 different purchases simultaneously how would it deal with that?

Think of it this way. You do not keep a copy of a bitcoin, or bitcoins. You cannot have the bitcoins themselves, even electronically.

A bitcoin is simply a balance that is held in a numbered account. An account to which you hold the key for. New accounts can be created at any time by anyone and you can have as many as you want. In practice, most bitcoin software will automatically merge all your accounts in to one "wallet".

What you keep as "your bitcoins" is not the bitcoins themselves, but the key that allows payments on the account in which they lie.

To spend some coins, you simply broadcast a message that says "Pay 1 BTC from account 1gGUI5t7GVVr57jkdujg679g to 1GBUJghilfyil8763bjklsdf. <Signed digital signature of 1gGUI5t7GVVr57jkdujg679g>"

Only the key that matches the account can create a valid digital signature.

So, for a transaction to succeeed. You need 2 things:
1. Bitcoins in the specified account
2. The key to that account.

Because the account ledger is public and distributed, it is easy to validate 1. Because it is easy to verify if a signature came from the legitimate holder of an account, but impossible to forge, it is easy to verify 2.

Notification of a transaction will "received" instantly. Network "confirmations" follow a timetable which operated at roughly 10 minute intervals.

If you attempt to overspend an account, then some transactions will fail to confirm. The ones that get into the network first will confirm, but once those confirm any overspend transactions will simply not confirm. Most software will "expire" transactions after 2 days if no confirmations are received.

There is a theoretical issue in the event of a severe network disruption, where part of the network might confirm a transaction, but when the network reforms, an overspend was attempted. In this case, "confirmations" can be reversed. However, this is exceedingly rare. (However, it could happen, if the all international UK internet connections were severed, and one transaction was made in the UK, and one in the US. Both transactions would confirm initially as they are unaware of the other transaction, but once internet connectivity is restored, one transaction will end up with its confirmations revoked (the one with the "smaller" network. In practice, this type of event is unlikely, and even if it did occur, transaction confirmations would likely be severely delayed in the "smaller" network).

Current recommendations is that 6 confirmations (about 1 hour) can be treated as absolute certainty. In practice, 1 confirmation (on average 5 minutes, but there is significant variability) would be sufficient for almost any transaction.

 

[Crusty]

Thanks for taking the time to answer my queations.



How quickly does the Bitcoin network register the change of ownership?

If I have 5 copies of my Bitcoin and click on 5 different purchases simultaneously how would it deal with that?

Having 5 'copies' of the datafile that is considered your wallet is not the same thing as duplicating 1 Bitcoin into 4 others. Your wallet file is just a datafile that contains a series of crytographic keys and some basic accounting information. The actual record of your coins is contained within the blockchain and locked with the use of the keys stored in your wallet file.

So in order to duplicate your coin into 4 more coins, you would have to fool the Bitcoin network into accepting a transaction that creates and gives you those Bitcoins, which it won't, because that's a malformed transaction.

As far as confirmation time, the Bitcoin network will start to confirm a transaction within the first 10-15 minutes if you paid an appropriate fee.

Once a transaction has been confirmed, the input coins for the transaction are spent and any further transactions that use those input coins will be considered invalid.

So even if you broadcast 5 transactions sending the same 1 bitcoin input to 5 different outputs, once the first of those 5 transactions becomes confirmed the others will be rejected because the input coins have already been spent by a confirmed transaction.

How can it do that with no personally identifiable information? Do you just mean the the identifiable information is encrypted or that it doesnt exist?
If its not registering who owns it I'm not understanding how its going to work out which is the "real" bitcoin and which is the "copy". Is it just the one thats used first? It would have to be a pretty fast network to combat any fraud that way wouldnt it?

I'm not trying to be difficult with these questions, I'm just having difficulty getting my head around how the "ownership" thing works.

This is a very simplified explanation, as there is a lot of nuance to how it all works, but basically...

All you need to do to receive bitcoins is create a public/private encryption key pair. Then anyone can use your public address to encrypt their transaction so that you are the only person who can decrypt said transaction(using your private key) and use the coins contained within it.

Usually, when you want to send money to someone you give them access to your private account through a debit card or a checkbook. Those authorize the recipient to retrieve money directly from your account. With Bitcoins, it works the other way. When you want to pay someone, you ask them where to send the money(their public key), and then using cryptography you transfer the ownership of coins to them. So you are pushing the payment to them, instead of them pulling the payment from you.

Every coin, and every transaction can be linked through this web of encryption keys. The only way it is possible for someone to outright steal your Bitcoins or reverse a transaction is if they got access your private encryption key or took control of more than 50% of the networks computing power.

So there is no reason to attach any personally identifiable information to your Bitcoin addresses if your goal is to remain pseudonymous and only transact in Bitcoins. However, as soon as you want to exchange fiat currency for Bitcoins, you will have to give up your information to the service provider you are using because they have to follow money laundering regulations.

 

[WelshBloke]

...

...

Thank you both very much.

 

[Eug]

People were considered stupid for believing in or investing in bitcoins, but not as much was made of the cardboard platform they were using called Mt Gox, which at this point it's revealed to have had the most risk. People had too much of a false sense of security from dealing with actual regulated brokerages.

Quite frankly, I have zero confidence in the other bitcoin holders as well. Maybe Mt. Gox was worse, but that doesn't make the others necessarily trustworthy.

 

[dighn]

Looks like China is buying these up like crazy. I would have thought Mt.Gox announcing that 850,000 coins were stolen would have a bigger impact on the BTC value. Seems to be totally unphased by the news. People must really feel the bitcoin protocol has something to offer.

China buying these up is not a good thing for BTC. Due to the lack of good investment channels in China, the wealthy (or just people with money to spare) will dump their money into anything that looks like it might give them good returns, anything from antiques to exotic types of wood to rocks. It'll just blow up the bubble even bigger while adding little to its actual function as a currency.

 

[Imp]

No, once you transfer a Bitcoin from your control and the Bitcoin network confirms it, you can no longer transfer those coins, even if you have 1000 copies of it.

Is the network like Torrent where everyone, presumably mining, chips in some computing?

Or is there a centralized server that has to be maintained by a centralized organization?

I honestly don't know.

 

[Crusty]

Is the network like Torrent where everyone, presumably mining, chips in some computing?

Or is there a centralized server that has to be maintained by a centralized organization?

I honestly don't know.

Yes, mining is a distributed process. No, there is no centralized server and there certainly is no requirement for it.

Mining is not necessarily peer-to-peer in the fashion that torrents are, but it is similar. Mining is only part of it though, there are also 'full Bitcoin nodes' which communicate with the miners to relay incoming transactions from the network to the miners and store the finalized block that is generated by a miner in the blockchain.

 

[Mark R]

Is the network like Torrent where everyone, presumably mining, chips in some computing?

Each individual node has various functions that it can (but does not have to) perform.

In practice, most nodes will need to hold a complete copy of the "blockchain" (the ledger of all transactions and accounts).

Nodes:
Generate transactions (e.g. a user wanting to spend their coins)
Relay transactions (e.g. an online user)
Confirm transactions (a miner).

Most nodes will, by default, relay any transactions that they receive (subject to firewall/router limitations).

Optionally, the node can contribute some computational power, in order to verify and confirm transactions. (This is what is termed mining)

In the case of mining, the process is essentially a lottery. A miner verifies all the broadcast transactions that they receive and compiles them into a ledger. They then keep trying random numbers until the "cryptographic hash" of the ledger + random number equals a pre-defined range of values. The only practical way to solve this problem is to keep trying number after number, and computing hash after hash, until you find a number which works.

Once a miner finds the number they transmit it, together with all the transactions that they have checked for validity, allowing it to be added to the blockchain (and results in the network issuing a "confirmation"). Once their "block" has been added, some new coins are spawned into the miner's account as a reward. The miner needs to check all the transactions for validity, because if another miner finds an invalid transaction, then when they submit their block, they can invalidate the first miner's reward, and claim it for themselves. This process helps keep miners honest.

 

[artvscommerce]

China buying these up is not a good thing for BTC. Due to the lack of good investment channels in China, the wealthy (or just people with money to spare) will dump their money into anything that looks like it might give them good returns, anything from antiques to exotic types of wood to rocks. It'll just blow up the bubble even bigger while adding little to its actual function as a currency.

I see your point, but I think peoples confidence will only help the adoption. I believe the first bitcoin retail store was just opened in China. Overstock.com and Tigerdirect have seen millions in bitcoin sales already. I agree that most of the current value is based on speculation, but there are still LOTS of business accepting it and lots of people spending it on goods and services. And this number seems to be growing quickly.

I think some of the reasons people use to point out how it will never work and never be accepted by business are quite funny. When business accept BTC, they can have it converted to USD immediately by their payment processor if they choose. So they do not need to worry about the value fluctuating. The other comment I often see if that the transactions can take hours. This is simply not the case. By adding a fee on for a few pennys, i've had all of my bitcoin purchases confirm in seconds. I made a purchase recently using BTC and the entire process was much quicker, easier, and more secure than using a credit card.

 

[BoberFett]

Cryptography is a topic makes my brain hurt, but anytime you have a distributed network and no central authority, isn't there an opportunity for rogue systems?

 

[artvscommerce]

Cryptography is a topic makes my brain hurt, but anytime you have a distributed network and no central authority, isn't there an opportunity for rogue systems?

I'm with you on brain hurt part, but depending on what you mean by rogue systems, I haven't seen any evidence of that yet. So far the thefts i've seen have been quite predictable.

 

[Crusty]

Cryptography is a topic makes my brain hurt, but anytime you have a distributed network and no central authority, isn't there an opportunity for rogue systems?

Yes there is, but if you want your rogue system to have any impact you're going to have to provide at least 50% of the networks computing power with your system.

That is not an easy feat.

 

[Mark R]

Cryptography is a topic makes my brain hurt, but anytime you have a distributed network and no central authority, isn't there an opportunity for rogue systems?

Yes there is. This is a problem known as "the byzantine generals problem".

The reason bitcoin is so "wasteful" of energy and computing resources is to try to deal with this problem.

Control is awarded democratically, effectively one hash, one vote. The only way an entity can exert full control is to hold 51% of the CPU power of the network. If you try to manipulate the network, but hold less than 50% of the CPU power, then you will be out voted and your effort will be wasted.

 

[_Rick_]

Yes there is, but if you want your rogue system to have any impact you're going to have to provide at least 50% of the networks computing power with your system.

That is not an easy feat.

ASICs made it more feasible though. These put mining well out of the reach of the general public, and even when not running in rogue-mode they generate some income. If someone has a few 10s of millions to put into ASIC miners, I'm pretty sure the market could be crashed, and massive gain could be made...

It might not even be illegal, after all, that's how the system is supposed to work.

 

[ImpulsE69]

Yes there is, but if you want your rogue system to have any impact you're going to have to provide at least 50% of the networks computing power with your system.

That is not an easy feat.

I think we had a conversation before revolving around this and everyone kept saying how it would take 1000's of computers millions of years to break the encryption. Ok, I get that, but the problem is that's not where you look when you want to get something. If enough skilled people want into something they will find a way. They just look for the weakness. There is too much shadiness around the whole thing and it is prone to people trying to take advantage of it. Has the evidence been shown that the Bitcoins from Mt. Gox are indeed gone, or did they just change hands and exchanged for $$ and hence disappeared from the overall mesh.

 

[Crusty]

ASICs made it more feasible though. These put mining well out of the reach of the general public, and even when not running in rogue-mode they generate some income. If someone has a few 10s of millions to put into ASIC miners, I'm pretty sure the market could be crashed, and massive gain could be made...

It might not even be illegal, after all, that's how the system is supposed to work.

I don't see how ASIC's made it any more feasible than it was before.

Even if you had all this money to put into ASIC's, you simply can't find the supply to buy enough hashing power.

What worries me, and is totally possible, is that someone finds a backdoor into some of the ASIC devices and is able to take control over large amounts of them. That would be a far more plausible attack on the Bitcoin network then simply trying to buy enough hashing power to try and out-compete everyone.

 

[Crusty]

I think we had a conversation before revolving around this and everyone kept saying how it would take 1000's of computers millions of years to break the encryption. Ok, I get that, but the problem is that's not where you look when you want to get something. If enough skilled people want into something they will find a way. They just look for the weakness. There is too much shadiness around the whole thing and it is prone to people trying to take advantage of it. Has the evidence been shown that the Bitcoins from Mt. Gox are indeed gone, or did they just change hands and exchanged for $$ and hence disappeared from the overall mesh.

No, there is no evidence that the Bitcoins were stolen out from under Mt. Gox.

In fact, Mt. Gox has been saying all along that the coins are 'temporarily unavailable'.

To me that means 1 of 3 things.
1) Technical problem. Failed RAID array waiting for a rebuild, etc.
2) They physically lost their private encryption keys
3) The feds have seized their coins and keys during their investigation of the silk road. The operator of the silk road had 10's of thousands of coins stored on Mt. Gox so it is not out of the realm of possibility that the feds seized all of Mt. Gox's assets until they can sort things out.

 

[_Rick_]

I think we had a conversation before revolving around this and everyone kept saying how it would take 1000's of computers millions of years to break the encryption. Ok, I get that, but the problem is that's not where you look when you want to get something. If enough skilled people want into something they will find a way. They just look for the weakness. There is too much shadiness around the whole thing and it is prone to people trying to take advantage of it. Has the evidence been shown that the Bitcoins from Mt. Gox are indeed gone, or did they just change hands and exchanged for $$ and hence disappeared from the overall mesh.

No, M:TGox just spent coins they kept twice, but internal book keeping only retained one transaction. The coins aren't gone, they're with the people that knew how to 'sploit M:TGox's broken "internal" accounting, and had them wired to their accounts.

As for rogues ASIC take-over: you would have to start a bit higher in the process, and basically make sure that Intel/TSMC are building the ASICs you designed. When they were offered to the public, the advantage shrunk slightly, and the cost of entry became a bit higher. If you're not in cahoots with the designers of the ASICs it's going to be much more difficult to get in on that. And it's probably not practical. But considering the potential windfall (you either break BTC and can make billions by betting against it or you have free transfers of however many bitcoins as you wish), even an initial investment of a 100 mil could pay itself back quite quickly.
Sadly, I'm too lazy to look at the actual numbers, and how much it would cost to replicate the current mining power, given your own ASIC design and fab contract.

 

[Crusty]

No, M:TGox just spent coins they kept twice, but internal book keeping only retained one transaction. The coins aren't gone, they're with the people that knew how to 'sploit M:TGox's broken "internal" accounting, and had them wired to their accounts.

As for rogues ASIC take-over: you would have to start a bit higher in the process, and basically make sure that Intel/TSMC are building the ASICs you designed. When they were offered to the public, the advantage shrunk slightly, and the cost of entry became a bit higher. If you're not in cahoots with the designers of the ASICs it's going to be much more difficult to get in on that. And it's probably not practical. But considering the potential windfall (you either break BTC and can make billions by betting against it or you have free transfers of however many bitcoins as you wish), even an initial investment of a 100 mil could pay itself back quite quickly.
Sadly, I'm too lazy to look at the actual numbers, and how much it would cost to replicate the current mining power, given your own ASIC design and fab contract.

I roughly calculated that if you were to buy the current highest power ASIC at MSRP you would be spending > $100m and < $1b to match the current 35,000,000 GH/s.

 

[Number1]

Bitcoin bank Flexcoin shuts down after $600,000 theft

 

[_Rick_]

"While the MtGox closure is unfortunate, we at Flexcoin have not lost anything," Flexcoin said via their twitter account as recently as February 25th.

Comedy gold

 

[Phoenix86]

Not that I think BTC is good idea or anything but what does a bank getting robbed have to do with a currency's viability?

 

[_Rick_]

Not that I think BTC is good idea or anything but what does a bank getting robbed have to do with a currency's viability?

With a normal currency you can at least hope to trace it. With Bitcoin, almost every theft is a perfect theft - and you can only sue for breaking into a computer system anyway.

 

[Crusty]

With a normal currency you can at least hope to trace it. With Bitcoin, almost every theft is a perfect theft - and you can only sue for breaking into a computer system anyway.

Please explain how it is a perfect theft.

 

[BoberFett]

With a normal currency you can at least hope to trace it. With Bitcoin, almost every theft is a perfect theft - and you can only sue for breaking into a computer system anyway.

And how would you do this trace with paper currency? Do you really believe that whole sequential bills thing that conveniently solves crimes on TV? That only applies to newly printed currency, not the large amounts of cash that move around from day to day.