How are you viewing it?

Exactly! If it gets too popular, nobody will want to use it. Just like restaurants that are so crowded you can't even get a table -- nobody goes to those.

If Ripple is widely adopted, there will be a huge incentive to running servers. People who do high-volume transactions, arbitrage, or the like will need a high-bandwidth pathway into the Ripple network, and nobody is obligated to provide it to them. If they don't run a server, they may have to move from server to server, get banned, be required to do proof of work to prove they're not malicious and so on. The best way to do it will be to run your own server and share the burden, then other servers will be willing to give you everything you want because you are also giving that to them.

Also, I think your analogy to Bitcoin isn't quite right. Bitcoin has to reward miners because it needs massive amount of computing to secure the block chain. It's hard to imagine people will just pile on more and more computing power for nothing. Ripple doesn't need proof of work to secure it. A closer analogy to a Ripple server would be a regular Bitcoin server that's relaying transaction and serving blocks to anyone who wants them. People don't get any reward in Bitcoin for relaying transactions or serving blocks. (This analogy isn't perfect. Ripple servers do more work than non-mining Bitcoin servers do but still not nearly as much as mining. And miners have an incentive to make sure blocks can move quickly because they need that to keep their block rewards.)

I don't expect this will actually be a problem. I think there are plenty of individuals, non-profit organizations, and commercial operations that will find it in their interest to run Ripple servers. If not, then it will only be because Ripple isn't providing enough value to justify the cost of running it, in which case it would deserve to fail.

We tried to find a way to do that, but we couldn't come up with one. There's no way to measure system-wide how much work someone is really doing as opposed to simulating fake servers that don't do any real work.

It is possible that servers might charge people for "priority" or "high volume" connections. I don't think that model will be necessary, but it's always there.

Thanks. Expect to see much more from us in this area. We haven't specifically tried to approach anyone but the Bitcoin community yet but we will.

I don't think FedEx is really either lying or mistaken, they're just providing a summary. Essentially, they're just warning you that you need to check the actual regulations before you export precious metals to the Czech Republic.

http://www.fedex.com/us/international/irc/profiles/irc_cz_profile.html

So your argument is essentially that so long as Coinabul's terms were knowable to the customer prior to the order being placed and the customer placed the order nonetheless, Coinabul cannot be scamming if they follow their own terms.

Coinabul's TOS says "If Coinabul is held liable to you for a transaction, liability will be limited to either the amount which you paid during the transaction or $100; whichever is greater."

So if I place an order for $1,000 worth of gold, Coinabul opts not to send me the gold for no reason, I sue them, get a $1,000 judgment and they pay me $100, that wouldn't be Coinabul scamming?

It is, unfortunately, fairly routine for companies to have terms of service that say "We can do whatever we want and your only recourse is to ask us nicely to do something else". While companies think this provides them with all kinds of legal protection, what it actually does is cause people to have no choice but to ignore the terms when determining if the company acted responsibly or not. Coinabul's terms are these kinds of terms -- many of them contain "we can do whatever we want and you have to take it " clauses, such as this beauty: "Coinabul is not liable for any errors, negligence, or inability to fulfill any orders."

I'm not saying Coinabul shouldn't have this kind of ToS. While I think it's idiotic, I understand that lots of companies do this, often as a result of specific legal advice. And I'm not a lawyer. But it forces people to basically ignore the Tos when deciding whether to buy from the company or evaluating whether the company acted reasonably.

The reason nobody bothers to read a ToS is because they all say the same thing -- we can do whatever we want.

And here's where it hurts you -- this would be a reasonable term of service. So if Coinabul could argue that its ToS was reasonable and that customers agreed that they were treated fairly so long as Coinabul complied with its ToS, I'd agree 100%. However, because Coinabul's ToS is of the "we can do whatever we want" type, "we complied with our ToS" can't provide a defense against an accusation of scamming. So "make your ToS make you liable for nothing" may be good legal advice, but it fails in the court of public opinion because it prevents you from being able to point to your ToS to justify your conduct. An unreasonable agreement can't justify anything.

You can set a fee for the transfer of your IOUs from one person to another. You can set a premium when you take someone else's IOUs. Charging interest breaks this model though, so we can't support it inside the normal flow of payments. You can, however, achieve the same thing with an issue premium (or a "quality in" in Ripple terms).

See this thread for details on why interest doesn't work:
https://ripple.com/forum/viewtopic.php?f=1&t=1224

The simple reason interest breaks this model: It makes it impossible to compare the value of payment paths. For example, say you get $10 at a gateway but in exchange you lose a $9 debt at 12% interest. Is that a good deal? Well, you can't say unless you know how long that debt would have continued to pay interest.

If you do know how long the debt will pay interest, there's no need for interest at all if you can charge an issue premium. There's a one-to-one correspondence between "$X at Y% interest paid in Z months" and "$A paid in Z months at no interest".

Also, because IOUs are powerful in Ripple, it's not clear who should charge who interest. As an intermediary, you exchange one IOU for another. Presumably you consider these IOUs equal in value (or you are getting a more valuable IOU for a less valuable IOU) otherwise you wouldn't make the deal. So why should anyone pay you interest after an even exchange?

Fedex's terms at http://www.fedex.com/us/international/irc/profiles/irc_cz_profile.html say, "An export license is also required when exporting toxic commodities, steel and steel products, precious metals, raw wood and wood products." Coinabul is the exporter, right?

That's the opposite reasoning not the same reasoning. If you're going to apply the same reasoning to the opposite situation, you get the opposite result, not the same result. You are arguing that the same reason applied to the opposite situation gives the same result, which it doesn't.

The reasoning is that no party has the opportunity to make a strategic decision to change the terms from what was agreed if they are better off with different ones at the expense of the other party. If a party makes the decision to cancel or seek a refund or whatever, this decision may not make them better off at the expense of the other party compared to what would happen if they went through with the sale as agreed.

I think you missed the point that Ripple is a credit and payment system that allows people to borrow and pay fiat currencies like dollars and Euros. You can compare Ripple and Bitcoin, and people inevitably do, but the main problems Ripple is intended to solve are totally different from the problems Bitcoin is intended to solve. A hammer isn't as good at putting in a screw as a screwdriver is, but it's what you need if you want to drive in a nail.

We plan to support it through contracts, but our preferred lending model uses an issue discount instead of interest. For a variety of reasons, that model works better (because of the nature of Ripple, not necessarily in general). This is discussed in more detail here: https://ripple.com/forum/viewtopic.php?f=1&t=1224

That is correct. It's implemented in the server and works on the network but there's no easy way to set or view it yet.

Really? So I can take orders to be delivered in two weeks and if Bitcoin doubles in price between when I take the order and when I ship, I can just refund them half the Bitcoins instead? This rule makes sense if the party not holding the Bitcoins asks for the refund, but it doesn't make sense when the party holding the Bitcoins decides to issue a refund. The point is to prohibit the party that asks for the refund from strategically choosing whether to issue a refund based on the price of BTC. Your rule, if applied to the party holding the Bitcoins, accomplishes the exact reverse of the purpose of the rule.

To be clear, I'm not accusing Coinabul of anything. But in this case, imagine if Coinabul was malicious. They simply hold the gold and the Bitcoins and wait a month or so stringing the customer along. If the gold drops in value, they ship the gold. If the Bitcoins go up in value, they refund the USD equivalent of the original Bitcoin price. If this rule were applied to the party holding the Bitcoins, it would enable malicious sellers to profit in precisely the way this rule, when applied correctly to buyers, is supposed to prevent. (I'm not saying Coinabul did or ever would do this. Just showing that this is a bad rule because it creates an exploit when its whole purpose is to close one.)

The correct rule would ensure that neither party has the opportunity to strategically choose to ask for a refund by ensuring that a refund doesn't enrich either party at the expense of the other over what going through with the sale would have done.

Coinabul converted to USD to minimize their risk should the customer cancel the sale or should they wind up keeping the proceeds. But unfortunately for them, this increases the risk should they need to cancel the sale. Someone necessarily has to take that risk, and I don't see why it should be the customer.

(By the way, I love Coinabul and prior to this incident highly recommended them. This is distressing though, and I'd strongly urge Coinabul to try to work out some kind of compromise that leaves the customer not significantly worse off.)

"I want to design a really nice suspension bridge. Now if this were your ideal suspension bridge, what would it be made out of? How many lanes would it have? Where should it be?"

"I want to make a really excellent book with high-quality binding, bright pages that are easy to write on and won't let your ink run, easy to hold. Now what should the story be about?"

You're running the design process in the wrong direction. Start with the problem you want to solve, not a list of solutions.

It doesn't do anything special. They just sit there, possibly forever.

It's theoretically possible that some means of recovering them could be added in the future, but it's not very likely. I proposed adding them into a mining reward fund and allowing each miner of a block to claim up to a certain amount from the fund if there was any left. But the last time this was discussed there was a consensus that you should be able to go to jail for 30 years and get back your Bitcoins from your brain wallet. I don't think changing it is a particularly good idea because we want people to be able to rely on Bitcoin's stability of its rules.

Wasn't one of the advantages of Bitcoin supposed to be that you don't need anyone's permission to use it? Rather than asking someone to stop or, worse, trying to make them stop, the better path is to fix yourself so they're not hurting you.

I'm not quite sure I follow you. You decide who you trust to hold money for you. If you don't trust someone to hold money for you, just drop their trust to zero and add trust for others.

Jed McCaleb has stated that Opencoin's business model is to hold XRP. XRP are intended for use in stopping spam (to be more precise, to judge the relative importance of transactions at times when transaction fees are scarce and to impose a "cost" on space in the ledger), but there are certainly other ways people could use them.

It makes the code you write using it a bug if it relies on it not having that property/feature!

It's actually about half of that. About half of those people shouldn't have found blocks and only did because the person who should have mined the block was working on the other chain.

Think about it this way, for each block in the fork, someone mined the block in the 0.7 chain and someone mined it in the 0.8 chain. About half the time, the right person got it (because the forks merged when the work in them was roughly equal), and they still have it. About half the time, the wrong person got it and the right person lost it in an orphaned block.

There's no sane way to enforce timestamps. Say someone mines a block that you think has a time three minutes in the future, what do you do? Do you ignore it? Do you deliberately set your time even further in the future and try to mine a block with a time you believe is wrong?