Nobody will be buying 400GH/s miners at $5,000, but they might be buying 2TH/s miners at $5,000.

No, not at all. ASIC companies will respond by making more powerful miners using better processes and better designs. The difficulty will keep going up.

Nonsense. The balancing factors will be that it is barely profitable or barely not profitable with the current generation of miners but highly profitable with the next generation of miners, incentivizing ASIC companies to build the next generation. The governing factor that keeps the spiral from going up forever is the total value of all mining, which will go down as the block reward drops but also vary based on the price of a Bitcoin.

It's not unusual for USPS to return no information on a tracking number until after the item is delivered. Of course, if the item does not exist, no information will ever be delivered.

Right, good points.

The two things you really want to know:

Is it limited to 5A out or does it add an extra .5A it takes from the downstream? If not, 10 erupters will be dicey.

Is it monolithic or chained? If it's chained, many of the ports won't work on an rPi.

It works the same as any payment. The pathfinding engine looks at what assets you have and what assets the recipient will accept and then it looks at all the liquidity available in the ledger (ripple paths and order books) and finds the cheapest path(s) to make the payment. The payment can even pull from multiple paths and multiple order books.

The payment interface requires you to select the recipient account (and destination tag if appropriate), destination currency, and destination amount. You can then pay with any currency you hold, provided you have a sufficient amount and there's sufficient liquidity available in the ledger. (There are a few exceptions. For example, you can't make a payment to yourself with the same source and destination currencies.) This is basically Ripple's primary feature.

That's a bug and starting Monday, fixing it will be my top priority. (It seems somehow the pathfinding engine selects on order book and the path selection engine checks a different order book to see how much money the path can deliver. I have good logs and reproducible test cases.)

The thing that got me was that they placed sales offers on Bitmit (with escrow) and then when people bought them, they told them to pay directly. At an absolute minimum, they're trying to ripoff Bitmit for the escrow fee, and far more likely they're trying to rip off the buyer.

When you place an offer, you have to specify the exact asset to be traded. If you have Snapswap USD, you have to place offers on the SnapSwap USD order book.

The easiest way is to make an XRP payment to yourself and pay with USD. Alternatively, you can trade on the Snapswap USD to/from XRP order book. At the moment, if I'm reading this right, you can actually get very slightly more XRP per Snapswap USD than per Bitcoin USD!

If you can privately send me your ripple address (the thing that starts with an 'r'), I can take a look at your account to make sure there's nothing weird about it. Otherwise, you can file a bug report at https://github.com/ripple/ripple-client/issues

I've never head of anything like either of your two issues.

That's really bizarre. Do you have some absurd number of offers on the books or trust paths already created?

You should just be able to enter "bitstamp" if Bitstamp is in your wallet already.

Paper gold has a risk cost and a convenience premium relative to physical gold. For most use cases, the convenience outweighs the risk. But it is a decision you have to make, and you must always keep in mind the fact that there is counter-party risk.

I'm assuming you mean BTC issued by Bitstamp.

The easiest way to do this is as follows:

1) In the Trust tab under Advanced, create a BTC pathway to Bitstamp. Set the limit higher than the amount of BTC you want to hold. (This essentially agrees that people, including you, can pay you BTC by making Bitstamp owe you BTC.)

2) Hit 'Send' and enter in your own Ripple address. Set the currency to BTC and enter the amount of BTC you want to buy.

3) If the transaction is possible, you will be given the option to pay with XRP. If the amount is acceptable, click the green button.

The more complex way to do this is as follows:

1) In Trade tab under Advanced, select BTC/XRP.

2) Click the Change Issuer button next to BTC and select Bitstamp.

3) At the bottom, click 'Order book' to see the offers available.

4) Click 'Buy BTC' and make an offer. If it completes immediately, you're done. Otherwise, it will stay on the order book.

The btctock.com website offers unusually low prices for Bitcoin mining hardware. The same person has traded on BitMit as 'btctock' and possibly trades elsewhere as well.

On at least two BitMit auctions, the seller attempted to convince buyers to cancel escrow and deliver the Bitcoins directly. Because the website promises low prices, claims immediate availability, only takes payment by Bitcoins, spells its own name wrong in its site graphics, and its owner has no real identity that I could find, I strongly suspect that it's an outright fraud.

The person I know who knows the most about the business side of Bitcoin shared an insight with me. A significant fraction of Bitcoins are held by miners. They are long-term bullish on Bitcoins, so they hold rather than selling, essentially hoarding Bitcoins and keeping the price high. Now, ASICs are about to totally change the mining landscape. They either have to give up on mining or buy ASICs. Since they're long-term bullish on Bitcoins, they believe it makes sense to buy ASICs and invest in mining. But to do that, they need cash. So they're selling their Bitcoins to get cash to buy ASICs. This is pushing the price down. Essentially, the value in the Bitcoin economy is being sucked out by ASIC manufacturers. If this person is right, we could see low prices at least until the vast majority of miners are on 28nm ASICs.

If you trust someone or something that is untrustworthy, you can lose to the extent of the trust you extended. This really shouldn't be surprising. If I trust you to owe me $50, I can lose $50 if you aren't trustworthy. (We plan to change the system mechanics a bit anyway.)

The hub being the problem would be my guess. Each USB erupter draws 510mA, assuming you have the newer ones. Typical 10-port hubs only ship with a 4A power supply, which would limit them to 7 erupters. Some can take up to 500mA from the host, but some hosts can't supply that and even if your can, that only gets up up to 8. Many 7-port hubs have a 3A power supply, which will get you 5 erupters. If removing erupters makes it work stably, it's likely the hub or power supply.

If you have a PC power supply handy, it can probably supply 5A at +5V off any of its power connectors without a problem. You can splice that to your USB hub's power cable.

I think for driving use and adoption of Bitcoin over the medium to long term, price stability is what you want. Increasing prices make hoarders happy and drive speculation, but if you want Bitcoins to be adopted as a means of exchange (the gold standard for broad adoption) you need a fairly stable value. If nothing else, rising prices drive fear of a sudden drop which stunts adoption as a means of exchange.

No, you atomically buy the right to those Bitcoins as part of the payment process. Someone has to have them to sell to you and you have to have something they're willing to trade for them.

My definition is similar. I define a pool as hoppable if there's some strategy a miner could use to intermittently mine in that pool such that he reduces the expected return of the other miners below what they would get if every miner stayed in the pool full time. That is, if the timing of other people's mining can reduce the expected return of a full-time miner, the pool is hoppable.

I don't think it's exploitable under most practical conditions. You'd need to control a significant fraction of the mining power in the pool and probably also have better latency to the pool than every other miner. Nevertheless, it's a poor choice of methods given that it's so easy to construct a method that's provably hop proof. Again, assuming I understand the payout method correctly, which seems very hard to believe given that the pool claims to have a payout method mathematically proven to be hop-proof and the method as explained on their web page clearly isn't.

They're not equally probable. The probability depends on latency to the pool and mining speed. All that's necessary to prove that a pool is not hop proof is to show that there exists some time an attacker can identify where the payout is greater than the payout from solo mining. The claim is that the payout method is mathematically proven to be hop proof. That seems impossible, again, unless I misunderstand the method.

All that's needed is for my latency and mining rate to be sufficiently good that the reward I expect for my first share if I don't find a block outweighs the risk that someone else will submit a share before I find my first share. If I understand this method correctly, it cannot be mathematically proven to be hop proof because you have to make assumptions about latency and mining rate.