Accounts don't have corresponding addresses on the send.  Addresses are linked to account for RECEIVING only (and only to determine which "account" the coins should be credited to. 


Likely a typo I assume you mean address.

If you send from account "A" then account "A" balance will be reduced.  Funds can be sent from any address in the wallet.
As Gavin indicated like in a modern bank (or shared wallet) the coins are combined.  Accounts merely exists as an accounting mechanism. 

What would be more useful is a "multi-wallet" capability.  The ability for a single wallet.dat to contain 2+ discrete wallets or alternatively the ability for a client to load 2+ discrete wallet.dats.

MtGox already offers bank wires.  There is no possible way to move funds faster in the fiat banking world.  MtGox payouts by wire take 14 to 21 days.  Kinda hard to see how the slower ACH/Direct Deposit would improve anything.

Nope.  Neither did the Dwolla.  My wire is still "confirmed" not "processed" after 14 days and my Dwolla is similar after 12 days.

Don't over think this.   If you owe someone $100 you simply give them $100 worth of Bitcoins.  If you owe someone $212.87 you give them $212.87 worth of Bitcoins.  

Also the offer my company made above still stands ...

Because in strong cryptography the algorithm should never be considered a secret.  The secret is the secret and it doesn't matter if the attacker knows everything else.  One should assume the attacker already knows everything but the secret.

A smart hacker can easily determine the algorithm anyways.  Simply locate a known account from the password dump. You have the plaintext & the hash.  Determining the algorithm is trivial.

For example, our site uses bcrypt (workload 10) with a random 128 bit per record salt.

Here is an example password hash (bcrypt includes salt and algorithm version in the hash output)

You have the algorithm and the salt; everything you need to crack the password. 

Still you will NEVER (not today, not before the sun burns out) brute force that password.  Hell if you do I'll pay you $1,000.

A double spend is when the attacker makes two transactions; one sending coins to you and another sending the same coins back to him.

The tx included in blockchain is the only one which matters.  That being said pulling off a double spend in meatspace is non-trivial.  Each node will ignore any double spend it encounters.  

So the attacker must:
a) ensure the victim sees the "good tx" of the pair first (or they will never see the good one).
b) ensure that a significant fraction of hashing power sees the "bad tx" of the pair first.
c) ensure the victim doesn't see the "bad tx" before the sale completes (just got busted double spending).

While it isn't impossible it is very difficult.  A cautious/smart merchant would run a modified verision of the bitcoin.  The merchant would establish a huge number of connections.  When it receives its "good tx" it would ignore it unless it came from a trusted node.  It also wouldn't relay the tx to any node.  By only accepting incoming tx from trusted nodes (major exchanges, major ewallets, major mining pools, other merchants, etc) the merchant would have some confidence that the "good tx" has been seen by a significant portion of the network.

This would greatly reduce the possibility that an attacker could accomplish a,b, c.  Bitcoin is somewhat unique in that fraud involves the attacker's own funds (unlike CC fraud where attacker has no "skin in the game").  That changes the risk vs reward balance.

It really depends on the amount of funds we are talking about.  Trading something in the real world to someone you know personally for a small amount of coins is totally different than trading 1,000+ BTC for irreversible funds online with an anonymous party.

Also the passphrase is converted into the AES key by doing ~10,000 SHA-256 operations.  This is done to make brute force attacks less effective.  Unless you remember a significant portion of the passphrase (i.e. remember passphrase as "This is my password" and it actually is "This is my p@ssword" it it computationally infeasible to break into your wallet.

Of course if you could brute force your own wallet, so could an attacker which would make encrypting it worthless to begin with.

Couple of counterpoints.

1) Markets trade on perception not reality.  If every single new channel, online blog, and tech site is gabbering nonstop about SHA-256 being fundamentally flawed it is going to affect the market.

2) If you think "Gavin" or any Bitcoin developer is going to patch the core algorithm of the network in 1 day you must be in some alternate universe.  Maybe within 1 month if the flaw is 0-day critical.  I would say more like 3-4 months if the flaw a threat but not immediately useful as an attack.  Have you seen how changes in Bitcoin protocol have taken in the past and these were more like tinkering around the edges.  Replacing SHA-256 with an alternate hash would be ripping the heart of the protocol out.

Nobody said "SHA-256 theoretical flaw = bitcoin insta-dies".  However to think it wouldn't be bearish on price is kinda short sighted.

USPS makes no guarantee as to delivery time.  If you want a guaranteed delivery service you need to use a guaranteed delivery service.  Second day Fedex is going to cost a lot more than $5.  For me it is $17.26 for someone with no FedEx account it is $26.21.  

Still even if you used FedEx if they fail to delivery on time the contract specifies maximum damages are the postage paid.  Lastly proving in court that someone lied is significantly harder than just showing they didn't do their job.   Can you prove beyond reasonable doubt that the delay was due to fraud and not simply a mistake, or error.  I doubt it.

As far as reporting the USPS to BBB?  LOLZ.  Think grandma is going to check the BBB before she mails in her cable bill?

USPS sucks and is cheap.  We ship out all our payment checks by USPS.  If a customer wants to we will ship by FedEx but it is going to cost a LOT more.  Express ("Overnight") USPS is $20 and I would say about 50% of the time does arrive next day and about 95% arrives within 2 days.  FedEx overnight is going to cost $46.95.

Mining is transaction verification.  It is the process of putting valid unconfirmed transactions into a permanent record.  The blockchain is that permanent record.  Miners hash unconfirmed transactions into blocks.  Mined blocks are distributed to every node in the network.  Nodes validate those blocks and add them to the end of the blockchain.  The blockchain is used to confirm the validity of future transactions.

Well while screw ups can happen most wires are automated now.  I would point out that in MtGox case wires take 18-21 days because MtGox doesn't do anything for 17+ days.   When they finally get around to submitting the wire to their originating bank the process is very fast; usually 1-2 business days depending on timing.

Until Amazon gets notified of the stolen credit card, reverse the value of gift code and makes Luceo account balance negative $10.  I give it 35 days.

wait the bank of dank needs an interest free loan?  say it isn't so.

0s or 0.1s, or 1.2s it doesn't really matter*.  When a miner solves a block he broadcasts it to his peers and those peers to their peers and those peers to their peers until every node knows about the new block.  From the period of time between a block being found and every node being aware of the block it is possible for another node to be unaware that block is already solved and two competing blocks to be produced.  If two miners broadcast competing blocks to the network before all the nodes are aware of the new block(s) then the network will temporarily be in conflict.  

If two blocks are found at the same time the same h  Until every node has been notified about a new block the potential for a race exists.  Some nodes will believe "a" is the correct next block and some will believe "b" is the correct next block.  The network resolves that conflict by the outcome of the next block.  If an "a" supported solves the next block then "a" becomes part of the longest chain and "b" is orphaned.  If an "b" supported solves the next block then "b" becomes part of the longest chain and "a" is orphaned.

Remember this is pretty rare only about 1% or so of blocks end up in a "race" with a competitor.  That critical window only lasts a few seconds.  Major pools an incentive to minimize orphaned blocks, because orphaned blocks produce no revenue.  The easiest way is to have a huge number of peers so within the latency of one connection they have already notified a significant fraction of the network.



* Technically the more of a head start one block has the more time it has to notify the network before the other block is found.  So the difference in time between the two blocks does affect the outcome of who wins the race but not the chance of a race itself. If 70% of the network hashing power considers "a" to be the next block and only 30% considers "b" to the next block then "a" has a 70% chance to win the race.

Because it is very hard to solve a block.
On average the entire network only solves one block every ten minutes.  This gives the network enough time to notify all nodes that a block has been found and when notified nodes beginning mining the next block.

However sometimes two miners/pools will solve a block at roughly the same time (within 3-5 seconds of each other) at which point the network is in conflict.  Eventually one of those chains will be extended another block and the competing block will be orphaned.

Example:
current block is #182800
Miner A & Miner B both solve block #182801
Miner A publishes it block to part of the network lets call it #182801-A.
Miner B publishes it block to part of the network lets call it #182801-B.
Each portion of the network mines to extend their version of the block chain. 
Say Miner C finds the next block (#182802) and it was using #182801-A as the prior block then #182801-B will be orphaned.

Direct Deposit is essentially ACH.  A flag is used in the NACHA file sent to the bank indicating the type of payment (recurring deposit) but other than that it is no different than any other ACH transaction.  Chargeback risk is the same as any other ACH.

Who would be stupid enough to deposit new funds there now?  I mean I know bitcoin users are often hard headed but I have to think the 5th hack/theft/fraud is finally enough for the idea to soak in.

About the only thing I would deposit there is some SolidCoins or flooz.  Maybe some Greek Drachmas or Zimbabwae dollars to support the fiat side.

I sent you a fraction of a coin.  Worth it for the lolz.

They are hoping it will cause the price to go down so they can buy more coinz!