You're correct. This makes it much harder to make an ASIC that has a huge mining advantage. I believe the thinking is that this prevents an evil conspiracy from buildings lots of super-ASICs and performing a 51% attack on the currency. But in reality, it allows an evil conspiracy to take the CPUs and GPUs they already have and use them to launch a 51% attack on the currency any time they want without having to invest in ASICs that can only be used to attack the currency and become worthless if the mining algorithm is changed.

A currency with an algorithm that can be accelerated by ASICs and that is being mined by ASICs is much more secure. An attacker would have to invest more than all miners have, and if they destroyed the currency, all their money would also be destroyed. This is where Bitcoin will be in a year or two.

rHzWtXTBrArrGoLDixQAgcSD2dBisM19fF

It's not true for information. There's all kinds of information that's scarce -- for example, tomorrow's stock prices or the factors of this near prime expressed in hex:

    00:c7:4d:7d:6e:dc:d1:c4:b1:8b:31:6b:0d:21:90:
    59:f6:f3:3e:9d:87:2b:04:de:99:39:4e:4d:a6:b1:
    bd:2a:55:25:ac:33:82:3e:68:19:1c:27:34:50:55:
    88:ba:9f:00:d0:5e:02:95:50:aa:d9:6a:67:62:ff:
    23:24:6d:d2:83

Say you hold the private key corresponding to a Bitcoin account that holds 1,500 BTC. Since that private key is just information and no information is scarce, you should be willing to give me that key for $1, right?

Nobody's figured out how to do this yet, it's a very hard problem to solve.

See this thread:
https://bitcointalk.org/index.php?topic=133096.msg1541453#msg1541453

They are always external. We considered the idea of a flag or indication in the ledger, but there didn't seem to be much point.

People, including myself, have been looking for a way to move Bitcoins out of the block chain and into another system (such that it can be moved back) without a central authority that holds the key needed to spend the bitcoins. So far as I know, nobody has figured out any way to do that. If you could though, it would be great for Bitcoin and make possible all kinds of new applications.

There's no good way to do this. Bitcoins could only be held on the Ripple system if someone were to gateway them, and there's no way to trust a particular gateway without having a central authority. Also, that would make Ripple's anti-spam system a Bitcoin spamming system as every Ripple transaction required a Bitcoin micro-transactions.

We honestly had no idea how to make that work for the reasons I describe above. Perhaps you could sketch out how your scheme would work.

Billions of XRP will be given away shortly, as the system is ready to scale.

I can give you a long list of reasons why that won't work. Maybe solutions could be found to each one of them, but I suspect that you don't see any reasons just because you haven't looked too closely at the problem. (But if you think you can make it work, by all means please work on it!)

If that happens, at that point Bitcoin will be broken and shouldn't be used until it's fixed. Fortunately, we should have plenty of warning.

But you don't turn the wheel of the car when you're 50 miles from the curve in the road.

Your argument is essentially that if members of a set are not scarce, then members of a subset of that set cannot be scarce. That is clearly false.

Yes.

Say, for example, an anti-spam scheme limited someone to ten emails a day. Would you want the ten email limit hidden so that you sent ten and then all of a sudden couldn't send any emails? Or would you prefer it be very clear that there's a ten email limit with a counter that goes down with each email sent? The way you stop spam is to make people cognizant of the resources they're using and aware that they are scarce.

Don't equate the internal system with the client. The client is designed to make it easy to do what we think people will want to do most. Right now, I don't think that's community credit, I think it's payment through gateways. Promoting community credit and making it easy for people to use and understand the consequences of what they're doing is a hard problem. I do truly believe that it could change the way people think about money in the future and having a payment system widely deployed that supported it will, I think, make it inevitable. But it's not Ripple's first step.

That's a client restriction. Internally, a currency is a 160-bit integer. One part of that namespace is mapped to three-letter currencies which also have a 16-bit 'version' (in case a currency is re-issued). Another part of that namespace is intended to be hashes of a "currency description". I believe custom currencies were discussed elsewhere and the basic idea is that development hasn't progressed much on them because of a lack of good use cases. The core of the system basically just treats them as 160-bit opaque values.

It is the nature of new technologies that business models have to change. Just ask the recording industry or the buggy whip manufacturers.

Suffice it to say that the exchanges we've been talking to haven't had trouble finding new business models that work with Ripple and, if Ripple succeeds, the old models might run into some trouble competing. If you don't think Ripple will take off, it's no threat. If you do think Ripple will take off, talk to us. There's lots of room and opportunity.

I think most of your complaints are legitimate. Every system reflects tradeoffs and Ripple is not trying to be all things to all people. But I have to disagree with this part:

The point is to allow people to transact in fiat currencies much the same way they transact in Bitcoins. Bitcoin is a currency with a built in payment system for that currency. Ripple is a Bitcoin-like payment system for any currency. "Ripple isn't a good currency" is a great rebuttal to an argument nobody's making.

Absolutely.

That would be a judgment call. You could definitely do it either way. If you want more centralized control than Ripple permits on the global network, you may wish to do it privately. If you want to take advantage of the global networks' (eventual) reach and in/out of cash, then the global network would be a better choice.

Handling large numbers of "currencies" and order books between them is exactly what Ripple is meant for.

Internally, currencies are just arbitrary 160-bit hashes and issuers are just 160-bit account IDs. The system was definitely designed to allow anyone to issue any IOU denominated in any currency if someone else was willing to accept it. The challenge is how to present that to users in a non-confusing and, ideally, intuitive way.

You are lucky. This practice is still illegal in my State (California) so I can't gain any of the benefits of this practice.

I doubt it, but you never know.

Ironically, that would probably be bad for Ripple in the short term. I think XRP would work best as a bridging currency if it had a stable value. And, of course, if there's a bubble that bursts, that would be terrible for use of XRP as a bridging currency. Fear of a burst can be as bad as a burst. And if there's no bubble, a burst is impossible.

I'm OpenCoin's (the company behind Ripple) Chief Cryptographer and one of the architects of the new Ripple. I'm not speaking for the company, only myself, but I guess I'm just the person saying the most. Thanks.

Because unless the Bitcoin economy is deep enough, such a transaction would incur a currency conversion fee on both sides. They'd have to buy Bitcoins and you'd have to sell Bitcoins. Of course if this works for your particular case, you should do it. And, of course, you can use Ripple to facilitate this. Say they want to pay with dollars but you're perfectly happy to accept Bitcoins. They can use Ripple to convert their dollars to Bitcoins right at the instant they pay you without paying someone commission on the currency exchange.

No, and I don't think that's necessary. All that's necessary is that they have some non-zero, stable value and that there be enough of them in circulation. I hope that will happen because it makes the Ripple system more efficient and everyone who holds XRP has an incentive to make that happen. But, of course, I can't promise it will happen nor can I predict the future.

Same person? I guess we don't have anyone with Satoshi's gift for explanation. Also, Ripple is much more complex than Bitcoin.

The proof is trivial if everyone holds IOUs from at least one major gateway, everyone trusts at least one major gateway, and there's a healthy path between any two major gateways. If you don't meet these conditions, it still might work.

Thus:

1) If you expect people to be able to pay you, you should accept IOUs from at least one major gateway in one of its major currencies.

2) If you expect to be able to pay people, you should hold IOUs from a major gateway in one of its major currencies.

3) Any time there isn't a healthy path between two major gateways major currencies, that's a profit opportunity for someone to provide such a path. Note that, by default, anyone who holds IOUs from one gateway and trusts another gateway (in the same currency) is a path between those two gateways. Cross-currency paths have to be explicitly offered by creating exchange offers.

It's very easy for one person to be an outbound island if they don't hold any IOUs that others trust. It's very easy for one person to be an inbound island if they don't accept any major IOUs that others hold. However, it's very hard to create a situation such that someone who is not an outbound island can't pay someone who is not an inbound island. To have two large disconnected groups, you need a case where nobody in that group has any path to any of the people in the other group. If XRP catches on as a bridging currency, of course, it will be all but impossible.

This will be especially true if gateways make it their business to ensure they have good paths to other major gateways. I don't know whether this will happen or not, but it seems like if it is a problem, a gateway would have a strong incentive to solve it.

I think there are a few ways that could work that might be interesting. For example, a minute after a block was found, mining pools could issue a signed statement that they commit to never publish a block not built on a chain that includes the recently found block. If you had such signed statements for significantly more than 50% of the hashing power, you wouldn't need to wait for any more confirmations.

But fundamentally, consensus can only work if you change the rules that the longest chain is the valid one. Otherwise, nobody can really promise a particular block will remain valid.

That would be a fairly fundamental change in Bitcoin.