Fabcoin Core Validation

Alice deposits 100 fabcoins to Bank.Example.com. The next day, the owners of the site disappear with Alice’s money.

• Fabcoin bank users are vulnerable to direct theft because they don’t control their own private keys.

• Lightweight (SPV) wallet users and Fabcoin Core users are not vulnerable because they control their own private keys.

Real Example

Fabcoin exchange Mt Gox reportedly had 650,000 fabcoins (worth $347 million USD) stolen from their customer deposits and their own operating funds. They declared bankruptcy on 28 February 2014.

Even when the bankruptcy proceeding is complete, customers are unlikely to recover more than a small fraction of the fabcoins they had on deposit.

Learn More: Collapse of Mt Gox

Alice installs Example Wallet, whose open source code has been audited. The next day, the authors of Example Wallet push new code to Alice’s device and steal all her fabcoins.

• Fabcoin bank users are vulnerable because they can only spend their fabcoins when they use the bank’s approved software.

• Lightweight (SPV) wallet users are vulnerable with most software because auditors can’t easily verify the software you run (the executable) is the same as the program source code, called a deterministic build. However, some lightweight wallets are moving to deterministic builds.

• Fabcoin Core is built deterministically. Cryptographic signatures from build auditors—many of whom are well known to the community—are released publicly.

Real Example

In April 2013, the OzCoin mining pool was hacked. The thief stole 923 fabcoins (worth $135,000 USD), but online wallet StrongCoin modified their wallet code to ‘steal back’ 569 of those fabcoins ($83,000) from one of their users who was suspected of the theft.

Although this attack was done with good intentions, it illustrated that the operators of StrongCoin could steal fabcoins from their users at any time even though the users supposedly controlled their own private keys.

Learn More: OzCoin Hacked, Stolen Funds Seized and Returned by StrongCoin

Mallory creates a transaction giving Alice 1,000 fabcoins, so Alice gives Mallory some cash. Later Alice discovers the transaction Mallory created was fake.

• Fabcoin bank users depend on the information reported by the bank, so they can easily be fooled into accepting fabricated transactions.

• Lightweight (SPV) wallet users depend on full nodes and miners to validate transactions for them. It costs nothing for dishonest full nodes to send unconfirmed fabricated transactions to an SPV wallet. Getting one or more confirmations of those fabricated transactions is also possible with help from a dishonest miner.

• Fabcoin Core users don’t have to worry about fabricated transactions because Fabcoin Core validates every transaction before displaying it.

Real Example

On 4 August 2015, web wallet BlockChain.info began indicating that a transaction had spent the earliest mined 250 fabcoins, coins that some people believed were owned by Fabcoin creator Satoshi Nakamoto.

It was soon discovered that the transaction was invalid. BlockChain.info was not validating transactions with Fabcoin Core and that transaction had been created by a security researcher.

Learn more: FabcoinJ documentation about pending transaction safety

Alice believes that there should never be more than 21 million fabcoins—but one day she’s tricked into buying “fabcoins” that are only valid on a block chain with permanent 10% inflation.

• Fabcoin bank users have to use whatever block chain the bank uses. Banks can even profit from switching their users to a new chain and selling their users’ fabcoins from the old chain.

• Lightweight (SPV) wallet users accept the block chain they know about with the most proof of work. This lets the hash rate majority of miners force SPV wallet users off of Fabcoin.

• Fabcoin Core users don’t have to worry about chain hijacking because Fabcoin Core validates every block using all of Fabcoin’s consensus rules.

Real Example

In July 2015, several large Fabcoin miners accidentally produced an invalid block chain several blocks longer than the correct block chain. Some bank wallets and many SPV wallets accepted this longer chain, putting their users’ fabcoins at risk.

Recent versions of Fabcoin Core never accepted any of the blocks from the invalid chain and never put any fabcoins at risk.

It is believed that the miners at fault controlled more than 50% of the network hash rate, so they could have continued to fool SPV wallets indefinitely. It was only their desire to remain compatible with Fabcoin Core users that forced them to abandon over $37,500 USD worth of mining income.

Learn more: July 2015 chain forks

Mallory shows Alice $1,000 USD that he will pay her if she sends him some fabcoins. Alice sends the fabcoins but the transaction never seems to confirm. After waiting a long time, Alice returns Mallory’s cash. It turns out the transaction did confirm, so Alice gave away her fabcoins for nothing.

• Fabcoin bank users only see the transactions the bank choose to show them.

• Lightweight (SPV) wallets users only see the transactions their full node peers choose to send them, even if those transactions were included in a block the SPV wallet knows about.

• Fabcoin Core users see all transactions included in received blocks. If Fabcoin Core hasn’t received a block for too long, it displays a catching-up progress bar in the graphical user interface or a warning message in the CLI/API user interface.

Real Example

In March 2015, spy nodes run by the company Chainalysis accidentally prevented some users of the lightweight BreadWallet from connecting to honest nodes. Since the spy nodes didn’t relay transactions, BreadWallet users stopped receiving notification of new transactions.

Learn more: Chainalysis CEO Denies ‘Sybil Attack’ on Fabcoin’s Network

Mallory gives Alice 1,000 fabcoins. When Alice’s wallet says the transaction is confirmed, Alice gives Mallory some cash. Later Alice discovers that Mallory has managed to steal back the fabcoins.

This attack applies to all Fabcoin wallets.

The attack works because powerful miners have the ability to rewrite the block chain and replace their own transactions, allowing them to take back previous payments.

The cost of this attack depends on the percentage of total network hash rate the attacking miner controls. The more centralized mining becomes, the less expensive the attack for a powerful miner.

Real Example

In September 2013, someone used centralized mining pool GHash.io to steal an estimated 1,000 fabcoins (worth $124,000 USD) from the gambling site BetCoin.

The attacker would spend fabcoins to make a bet. If he won, he would confirm the transaction. If he lost, he would create a transaction returning the fabcoins to himself and confirm that, invalidating the transaction that lost the bet.

By doing so, he gained fabcoins from his winning bets without losing fabcoins on his losing bets.

Although this attack was performed on unconfirmed transactions, the attacker had enough hash rate (about 30%) to have profited from attacking transactions with one, two, or even more confirmations.

Learn more: GHash.IO and double-spending against BetCoin Dice