I haven't fully thought this through, so there are probably holes in it. I have a mental picture of what a lightweight wallet client for a smartphone might look like. For the sake of argument, let's say that this client has some way to not require a full time connection to stay in sync with the full block chain. That in itself is a substantial problem to solve, but I'll gloss over it because it's not central to the point I'll make in a moment. Let's further stipulate that this lightweight client has its own wallet in order to carry around a small amount of spending money. Most of the user's savings are in a separate wallet, probably secured back at home, and possibly implemented as we are discussing here.

It seems to me that barcodes such as QR codes would be a natural way to transfer numbers between two devices in person. If you and I both have these hypothetical smart phone apps, and I want to give some BTC to you, then you could display your receiving address as a QR code, I scan it with my phone's camera, and then my client initiates a new BTC transaction using whatever method it has to do that. I like the idea of exchanging addresses via barcodes better than NFC, bluetooth, wifi, etc. because it requires deliberate user action, and it is easy to see that it is going on. In contrast, an attack over Bluetooth could go on silently without the user's knowledge. There may be a potential shoulder surfing problem... Needs more thought. A vendor might display their receiving address on a poster or display, depending on whether they want to use a different address for each transaction.

So, if a standard is developed for using QR codes to communicate between mobile clients, then maybe the same mechanism could be used to interface with a dedicated wallet device? I do like the idea of a purpose-built device that handles a wallet and does nothing else. While an existing device like an off the shelf Android device could be reprogrammed to serve this purpose, I'd feel safer with a device that has no network connectivity at all (not even hardware that normally is not turned on), in order to reduce potential remote attack vectors. There may be some security flaw that lets the device get owned by showing it malicious barcodes, but it would be hard to invoke if the device has a physical cover over its camera and it is locked in a drawer or safe.

If I understand correctly, on that OpenPGP smartcard the private keys never leave the smartcard. Instead, any signing with the private key happens on the smartcard. If the same idea was used for a Bitcoin wallet, then the private keys of the wallet would never get uploaded to the computer. Malicious code might still be able to use the smartcard to send out bitcoins from the victim's account while the card is plugged in, but it shouldn't be able to steal the private keys for later use.

The OpenPGP smartcard needs to receive a passcode before it'll do stuff. If the card reader has it's own keypad for entering the passcode (some readers do have such a keypad), that can help reduce exposure to malicious software since the passcode never passes through the potentially compromised computer.

I'll poke around for other threads on this topic, too...

I think that is technically feasible. There is a similar smart card for use with PGP encryption. I fact, it doesn't even need to be as big as a credit card. It could be the same size as a SIM card, and the PGP smart card is available in both sizes.

http://en.wikipedia.org/wiki/OpenPGP_card

How about allowing connection via HTTPS?

RandyFolds, If you think that it's a bad idea, then don't participate. Whether you choose to participate or not, skip the hostility. Calling the idea "fucking retarded" and calling other forum members "asshole" isn't helpful, even if it is true in your opinion.

If I use a Miley Cyrus Bitcoin Card to buy some smack on Silk Road... uh, I can't think of a good ending for the joke. Somebody pick up the ball and run with it, ok?

I think that'll make it hard to get UL listing for them. 

Neither the paper wallet service nor the wallet.dat protection service is useful to me, since I have the technical savvy to generate a wallet.dat on my own air-gapped computer, stick it in my safe, surround it with my dogs, and guard it with my own shotgun. I could even wear overalls and holler "git off mah land!" while I did it, although I'd have to buy some overalls (anybody selling overalls for BTC?). I say all this merely to establish my credentials as an uninvolved outside observer, posting my outside observations from my easy chair on Saturday evening because I can't think of anything more interesting to do at the moment. I can see how either service might appeal to somebody less well-endowed with technical savvy, safes, dogs or shotguns than I am. Different strokes for different folks, and all that.

Ok, holograms and fingerprints are neat, but at best they help verify that the postman (excuse me: postperson) didn't secretly copy down the wallet.dat before delivering it to me. I have nothing but the vendor's fine word that the wallet.dat was generated on an air-gapped computer. It's reassuring that he has offered up his ass for kicking in case of a scam, but I also have only his fine word that the name and address provided are really his own, and not those of a random stranger, or perhaps an ex boss whose ass is chronically under-kicked. PGP keys? Yeah, I can make those, too. They prove that all of the communications I receive that are signed with them originated from somebody who has access to a particular secret key; they don't prove the real identity of anybody unless I vet them somehow, such as by showing up unannounced at the provided address and having whoever is there offer me the matching key fingerprint, and perhaps a nice mint julep.

Don't worry, I'm not showing up at anybody's doorstep and demanding PGP fingerprints and refreshing beverages. And no, that's not a cauldron of boiling tar in the back of my pickup truck; it's an old desk that my employer was happy for me to haul away a couple weeks ago, and I've been too lazy to unload it yet.

My point is just that as a random dude on the Intarwebz, and absent any actual evidence of naughtiness, I see no difference at all in my perceived risk of fraud from either of the vendors in question in this thread. Some words on the screen of my iPad tell me that one stranger promises to make my wallet on an air-gapped computer rather than a PC at the public library, and some other words on the screen of my iPad tell me that another stranger promises to lock my USB thumb drive in his safe rather than hanging it from the rear-view mirror of his convertible next to the dream catcher. Unless there is some real evidence of a scam going on, neither party seems to have any basis to call the other a likely scammer. I just don't see any fundamental difference in the trust model of either service under discussion here, as perceived by a person who doesn't personally know either of the vendors.

Crappy little ATMs in gas stations turn profit the same way: by charging an exorbitant fee in exchange for getting money Right Now. A gas station ATM might charge $2.50 for a $20 withdrawal.

I agree that it makes little to no sense to stick a BTC kiosk into a random gas station yet, but a brick and mortar shop that chooses to accept BTC might also choose to stick a BTC purchasing kiosk in their lobby.

This seems to me like a concept that could be tried out at a small number of locations (even a single location) for prototyping and proof of concept purposes, and then naturally scale up with BTC acceptance.

Running a squadron of armored cars to service the machines would indeed be expensive. However, unlike ATMs that need to be loaded up with new cash, these machines would only need cash taken out. The machine could be bought by the proprietor of the store where it's installed, and then it's owner can periodically unload the deposited cash and buy more BTC to be "dispensed" on some suitable exchange. The machine sellers make their money by selling the machine, and the store owners make their money by selling BTC at a markup. No army of armored truck drivers is needed, because the machine owner teaks out the money and keeps it.

The BTC markup might be a flat fee, which people are already accustomed to at ATMs. Or it might be a percentage. It might be an explicit fee, or it might be hidden in the machine's cash to BTC exchange rate. That exchange rate might be based on live market data, or it might be adjusted less frequently with a proportionately larger markup to cover market variations. Those are implementation details that would need to be worked out, and different machine vendors/owners might take different approaches. The core of the idea remains that it would provide an easier way for a random person to turn cash into BTC, and the machine vendors would presumably provide a product and/or service to make it easier for a random store owner to sell BTC without being an expert in the field.

One of the neat things about this idea is that it could be tried out on a vey small scale... even at a single BTC-accepting business. Compare this to something like ASIC mining hardware (another topic that interests me): Both ideas make sense as things that will be nearly inevitable if and when BTC use grows the way that we want and expect it to. However, the up-front development cost to put the first BTC-selling kiosk in place should be at least an order of magnitude lower than the cost of making that first mining ASIC.

Regarding BTC accepting junk food machines, the biggest problem I see to be solved is how to make sure the user really pays for their candy bar without making them wait 10 minutes for a confirmation before the bar gets dispensed. That's a general hurdle for BTC use that needs to be solved well. Once that problem is solved, I can easily imagine BTC-accepting vending machines sprouting up in places like college campuses, particularly around EE/CS buildings or in dorms at technical schools. Some enterprising student at a place like Caltech or MIT will surely make a BTC-accepting vending machine for their dorm, if it hasn't already been done.

I am not presently a TH customer, but that latest response pleases me!

It seems to me that one of the great advantages of the near-instantaneous worldwide communication that we now enjoy on the Internet is that it is much, much harder to keep a scam going against people who make an effort to educate themselves. Scams now either prey on the utterly ignorant, or they're one-shot deals. As you point out, if you were to try scamming anybody on this board, there's already a pot of boiling tar and a sack of feathers at the ready!

Uh, by that reckoning, if you had started the process when you created this thread, you'd probably already have your USD... 

Because once all 21e6 BTC are mined, we'll all stop promoting and using BTC? 

I thought that I heard rumblings about MtGox dropping Dwolla while skimming the TH/Dwolla threads. Might need to hurry before that path is closed?

I just got a mental image of bums huddling around a BTC kiosk at night for warmth. 

Maybe this is a dumb question, since I'm not presently a user of either Tradehill or Paxum so I don't know all of the details. If you don't want to have to open a Paxum account, then couldn't you trade your leftover balance into BTC and then withdraw that to your wallet?

In terms of hardware, I think that the only difference between your proposed kiosk that you flash a QR code at vs. any other network-connected cash-accepting kiosk would be the need for an appropriate camera to look at the QR code... in other words, not much difference! The camera might double as a security camera. Maybe some kiosk designs already have a suitable camera in place for security?

Even though I've been critical of your posts in this thread, I think I agree with you on that point. There are many reasons to prefer not to do business with checks, but it seems to me that cutting checks to customers as a temporary stop-gap measure in this extraordinary situation is appropriate.

Good point. Handing control of your money to somebody else requires trust, and trust is fundamentally given with the understanding that an ass-kicking (whether literal or figurative) may result if it's violated.

They took action to plug a fraud hole that had already cost them tens of thousands of dollars. They also promised to spend their own money to make sure that their customers won't directly lose any of that money. As a result, some people are temporarily inconvenienced. Get over it.