FPGA hardware costs are more per Mhash/s than those for GPU hardware. 

Some people get electricity at either a tremendously low rate (e.g., those near hydro generation might pay as little as $0.05 per kWh or less even) or there is no marginal cost for consumption (e.g., included in the lease of an apartment or other facility).  For them, GPU is more profitable than FPGA due to the difference in price for the hardware.

Now BFL's ASIC changes the game though.   The cost of the hardware per Mhash/s is so much lower than GPU (under the assumption that BFL can actually deliver these and at the prices they've listed for sale).  Each $100 of BFL's ASIC produces as many hashes about at much as $1,000 worth of GPU hardware.   Thus the skyrocketing difficulty from even a marginal amount of ASICs delivered will make GPU mining yields so trivially low as to not be worth the effort.

Adding capacity now with GPUs hits physical limits, ... removing excess heat and access to electric circuits with enough amps.  Using FPGAs though allows a greater level of Ghash/s without hitting those heat and power problems for most.  Those might continue to at least be worth left running once ASICs arrive, at least for a little while longer.

The one-two punch to knock out GPU mining would be volume ASIC shipments and the arrival of the block reward drop (to 25 BTC, expected in December).

I see that LastPass has a way to view history, which if that showed login from an unknown IP address, that would be a pretty good clue.

I just tried to view the history but the LastPass UI for the date picker is so horrible I could not use it successfully.  (Top-right is the Lastpass asterisk (starfish, ironically   ) , then click History)

If I'm not mistaken, this status has persisted for several months now, either continuously or only sporadically where sufficient funds were available for them to accept Mt. Gox USD redeemable codes.

Presumably, this is because delays for withdrawal of USD from Mt. Gox through the banking system (either wires, or Dwolla) are so long that not only is AurumXChange not able to accommodate the redemption of MtGox Redeemable Codes at normal levels, it definitely cannot accommodate the influx when everyone else with funds stuck at Mt. Gox want to redeem at AurumXChange as well.

Lots of old coins moved recently.  Possible large withdrawals required coins fro Mt. Gox cold storage, which were a year or more old?



 - http://blockchain.info/charts/bitcoin-days-destroyed-min-year

If those eventually make it to other exchanges for sale, or back to Mt. Gox after a little washing, that could be some selling pressure.

"Hypothetically" the primary reason one would consider not withdrawing fiat is because the account has someone else's name on it.  But with Mt. Gox having delays with their withdrawals, it is conceivable someone would want to want to convert to bitcoin, move the funds to another place where they can be cashed-out.

So, the first question... does the exchange support redeemable codes or account-to-account (A2A) transfers of USDs? 

Mt. Gox,  Crypto X Change, Bitstamp, BtcTree and BTC-E all offer redeemable codes of USDs.
 - http://en.bitcoin.it/wiki/Redeemable_code

If so, then consider doing a trade by finding  a counterparty that wants to move their cash to the exchange that you use.  You exchange your redeemable code for your counterparty's cash -- whether it be Dwolla, or Western Union, or whatever. 

If redeemable codes are not offered, then you can try another widely used other approach.  Generally if you announce the size of your intended order in advance, you might find a counterparty willing to put up an opposing offer at a cross price that works for both of you.     You'll occasionally see this with BitFloor or Camp BX where someone has a large order that won't clear without pushing the pice significantly one way or the other.  But after pre-announcing that an order would be placed, by the time one is opened there is sufficient interest in unloading a batch of bitcoins for your slightly above spot price.

You had 40K BTC / $40K USD or more in Mt. Gox and weren't using a YubiKey or TOPT/Google Authenticator?  Seriously?

A couple of questions and suggestions.

1.) When I was going to give a prop, I was assuming I would enter the amount to give.  Only after clicking the Give Prop button did I see that this goes Flattr-style where all the props are dropped into the monthly bucket and then divided equally on my next billing cycle. If

that were explained at the top of the response page after I click "Give Props", then I as a new user would understand the action that just happened.

2.) Adjustments to "Needs Funds" props.  As I am impatient and didn't want to wait until the end of the month, I clicked "Divide the pie now".  I ended up with some showing "Needs Funds" because I had a higher pie allocation than I had a balance.

I'm left with Props with the status "Needs Funds".  I might have wanted to reduce my monthly pie amount rather than end up with unpaid props.  As now, when I do add more funds they go to pay the props that had been at "needs funds" rather than all props equally like I was thinking it would.

3.) No Unprop for "In Progress" nor "Needs Funds". Are these props permanent now?   What if I changed my mind, or Propped the wrong entry before the billing cycle executes?  Can't I unprop?

4.) No explicit amount.
I'm wasn't a fan of Flattr because I didn't really want to Flattr evenly.  I get Flattr's logic -- they determined that the subscription model was all that would work for them.  Well, that's what might work for them but the subscription model sure didn't work for me and I quit using them after just one month.

So instead of me sporadically giving some blogger $2 after or whatever after a great read -- say once every other month or so, I now instead Flattr exactly $0.00, exactly zero times -- but I do this each and every month now!

So do the math. ... $0 per month from people like me when the approach is to force a subscription, or $1 per month average if you let the individual choose how much and when to Flattr.

Ok, .. so, what I'm asking is ... must you follow the same approach as Flattr, or can you at least give me the option to do a one-time prop where I set the amount and then click "Prop Once" or something like that?

I know I can send using the anonymous Prop bitcoin transaction, but hey -- maybe I feel good about myself when seeing my name appear, or whatever, and would like this one-time prop to appear just like all my other props are shown.

Looking forward to the LocalBitcoins mobile app.

A few months ago was a StartupBus competition, and one of the contestants was Open Wallet app:



Here's their proof of concept video:

 - http://www.youtube.com/watch?v=QoRlyPZ3zJ0

I have no idea what they were going to use to settle payment to the party acting as "ATM".  I wouldn't doubt they had no good solution ... and left it as "and magic happens here", after figuring out that PayPal wouldn't work for this.

If that was the hurdle these entrepreneurs couldn't solve, then had they considered using bitcoin as the settlement currency they might have been able to get funding, develop it and launch.

Well, ... will LocalBitcoins take it further?

Paging Coinapult ...   

 - http://techcrunch.com/2012/07/12/power-to-the-cloud-twilio-takes-its-sms-api-global-adds-dozens-of-new-languages-150-countries

MOAR GLOBAL please!

Each jurisdiction can have different regulations on escrow.

Perhaps an escrow that deferred arbitration when the two parties don't agree to something like Judge.me would be viable:
 - http://bitcointalk.org/index.php?topic=83981.0

Yup.  In the U.S., some states require sales tax to be charged when buying locally, for instance.

 - http://thecoinologist.com/sales-tax-state-by-state-breakdown/





 - http://gold.bullionvault.com/How/GoldBullion/

In question is how much premium you pay to acquire the bullion. 4% to 8% seems pretty low even.

Just to put things in perspective, here's what the OP might have been looking at exactly ten months ago:



 - http://bitcoincharts.com/charts/mtgoxUSD#rg60zigDailyzczsg2011-06-09zeg2011-09-10ztgSzm1g10zm2g25

And as a store of value, yes -- bitcoin was failing for most people as at the time, most of those who held bitcoins by that time were underwater versus what they paid for their coins.

Could bitcoin see another drop like that, perhaps losing half its value or more?

It could.  If it happens it probably wouldn't be for the same underlying reason.  

It would have to be something that would cause a huge loss of confidence again, but it probably wouldn't be because of security at a single exchange.  Maybe in the U.S. FinCEN will put out a statement that Bitcoins are the same as "prepaid access" used internationally and that you cannot sell even 1 bitcoin without obtaining the identity of the customer.  Maybe an evil ASIC engineer will be first to come up with a couple dozen Thash/s of capacity and successfully double spend transactions that were previously confirmed six or more confirmations back.  Maybe the last patch Tuesday from Microsoft included a few lines of code which will, at a future date, send the Bitcoin/wallet.dat to the rogue employee's target.

We just don't know what happens in the future.

What we do know:

Bitcoin's security is better today than it was ten months ago.  Bitcoin's ecosystem then was not anywhere near as developed as it is today.

What we can presume:

Bitcoin's security will be better ten months from now than it is today (e.g., functional multisignature transactions).  Bitcoin's ecosystem will be much more developed ten months from now than it is today.

And the fun continues.   450K usernames and passwords from Yahoo! Voice:

 - http://www.trustedsec.com/july-2012/yahoo-voice-website-breached-400000-compromised/

There is a new user stuck in newbie forum jail asking to purchase these, so I'm cc:g the URL here:

 - http://bitcointalk.org/index.php?topic=92818.0

I see it is responding now, but all items are priced at 0.

So it might be coming back, apparently.  Monitoring this thread.

The bots can emulate anything the Javascript and browser would do, so if there is any financial gain from this then the bots will be mining you for income.  You couldn't do this, I wouldn't think, unless it was delivered as an executable and had anti-cheat smarts.

 - http://thejeshgn.com/2012/06/11/pyg2fa-python-library-google-authenticator-with-web-app/

One time passwords (OTP) are useful for protecting against replay attacks, such as what might occur on a compromised system that has a malware keylogger.

But to support this, the Bitcoin.org client would need to know the key.  If that key were stored on the filesystem or in the database it would need to be stored unencrypted.  So if the attacker has access to the database then the attacker has access to the key.   If the key weren't stored, and instead the user were prompted for the key, that would be something vulnerable to a replay attack.  So you really don't get much benefit from adding OTP to the Bitcoin.org client.

Now, a Yubikey can still be useful with the Bitcoin.org client though.  Although this has nothing to do with two-factor authentication, I see the Yubikey also supports a static password capability.   So if I understand Yubikey's documentation correctly, you could use the Yubikey in this static password mode when using the BItcoin.org client's wallet encryption.  You simply have Yubikey provide the passphrase used to encrypt the keys and then use the Yubikey each time the client asks for the passphrase (e.g., to add a new address or to send a payment).    I see the Yubikey support dual mode capability meaning you can use it for both a TOTP purpose and for a static password purpose as well.  I don't know if the Yubikey's from Mt. Gox still have this capability as those are modified Yubikeys.

Problem solved:
 - https://www.spendbitcoins.com/convert/ebay

As far as selling bitcoins for Dwolla, two market exchanges allow Dwolla withdrawal, generally without delay.

 - http://www.CampBX.com
 - http://www.Intersango.com

Another, FastCash4Bitcoins.com does the same type of Dwolla cash-out, however their funding levels are sporadic and today they have low Dwolla funds.

Legitimately, these would need to be "burned in" by mining.  So they can do a little of both -- mine some for themselves (while difficulty is low).

Watch for shipments to be timed right at each difficulty adjustment.  (e.g., mine before the difficulty change, then ship, then start the prep for the next batch for shipment 2,106 blocks later).

But they probably could never get the investment funding necessary to design and build these if they weren't pre-selling.  So even though this route does leave potential profit on the table (from not mining themselves), it does allow them to not just remain in the mining hardware vendor space but to remain the leader.

It depends.  What problem you are trying to solve?  If you are selling bitcoins, you want the best price.   If you are accepting Dwolla, then that person with Dwolla can buy bitcoin at Camp BX and Intersango at a little lower than Mt. Gox's BTC/USD level.  So if you set a price based on Mt. Gox's last, you might not be competitive and might lost the opportunity to do the trade.

These exchanges should trade at roughly similar levels because of arbitrage occurring.  But because of delays when withdrawing from Mt. Gox, there is a price premium when buying on Mt. Gox's BTC/USD market.

For historical data analysis, having exact values aren't critical and the trends over time will be the same.  (e.g., if the price premium when buying at Mt. Gox's BTC/USD is about 2% higher than at other exchanges, then it will likely still be 2% higher a month from now as well.)  So there is no burning desire for an accurate "current price" that draws from all the various exchanges.

The current price is of limited use anyway if larger trading volumes are being considered.  There isn't a terrifically large amount of buying or selling needed to cause the price to move a couple percent either way.  Only what sits in the order books will tell at any point in time what is needed for any trades in the several thousand dollar (or greater) range.