According to bitcoincharts.com, the Mtgox all-time high is $31.9099. Any minute now...

This has multiple use cases (some are explained at http://bitcoin.stackexchange.com/questions/3718/what-are-multi-signature-transactions), but software support is still in its infancy (http://bitcoin.stackexchange.com/questions/3557/how-does-one-make-multisignature-transactions-with-the-bitcoin-client).

Um, no.

Dollars are not scarce, the Federal reserve can print as much as it wants. (OpenCoin didn't "promise" not to print more, it is impossible by protocol)
Dollars are not durable, they can be easily lost and stolen.
Dollars are not portable, they need to be physically carried around, and in large quantities have significant physical volume and mass. (I'm assuming physical notes, if you consider bank balances as dollars they are slightly more portable but less scarce and durable.)
Dollars are not as divisible as Bitcoin / Ripple credits.

Happy Purim!

A few comments...

XRP as a currency.

XRP are not "just" what you use to send Ripple transactions. They are also the only thing you can hold in Ripple which is not an IOU. I think XRP will most definitely be a currency, with a non-negligible total market cap which floats against other currencies. And to understand why you need to look no further than why Bitcoin is/will be a valuable currency.

Bitcoins are valuable because they are in use, and they are in use because they are scarce, durable, portable, fungible and divisible; that is, a good currency. So are Ripple credits:

Scarce - There's an upper bound of 100B XRP.
Durable - You can store credits as private keys, back them up and encrypt them.
Portable - You can send them easily anywhere using the Ripple system.
Fungible - I don't know enough about Ripple but I'll guess all credits are the same.
Divisible - There are a total of 100 quadrillion drops in the system.

XRPs have all it takes to become a good currency, and I believe it will be used as such. A currency also needs to be current, and here psychological factors play a large role - the fact that many of the early adopters don't perceive it as a currency might slow down the process, but it won't stop it.

The fact that, in addition, XRP are what you use to send transactions in Ripple, will help them gain value as Ripple becomes in greater demand.

OpenCoin distribution model.

That said, there are issues with the distribution model that will either:
1. Enrich the creators above and beyond their fair share for their contribution, or
2. Slow down the adoption as a currency, as people are wary of #1.

In Bitcoin, the distribution model was democratic from the start. Anyone who would listen was able to take a part in the distribution and be equal to the creator and everyone else. Distribution is based on computational work which is measurable and unfakable. People have profited purely based on their belief in Bitcoin and their foresight to understand it. Knowing this is what gives me the power to overcome my envy for those who came before me.

But with Ripple, the creators get all the credits to begin with (some personally, some as part of OpenCoin). This means that the "creators are enriched by people using the system" factor is much more pronounced.

As far as OpenCoin's distribution plan - the fact that they have a plan, that is, that a centralized body is deciding where the credits should initially go to, is a problem in itself for a decentralized system. Also, I doubt a good plan can be come up with.

It will also be difficult for OpenCoin to prove they're not faking anything and "distributing" the credits back to themselves.

I don't think you can do much better than proof of work. You can use consensus for transaction synchronization, and a Bitcoinesque inflation schedule and PoW-based distribution (it would be like mining blocks without any transactions). If it was this way from the start Ripple could have truly been said to be decentralized.

That's good for you, but if I'm going to use the system, I do care what will happen to it after the flood. That's the point - with all due respect to the founders' business model, we are questioning the ability of this system to act in a sustainable, decentralized way.

Colored coins allow issuing assets without having to rely on 3rd parties.

You are presenting a false dichotomy between "payments are transactions in the blockchain" and "payments are processed by traditional service providers".

Bitcoin is a powerful technology which allows more advanced applications.

If you make multiple payments to a given merchant you can use payment channels. And if you want more flexibility you can use my add a third party to the mix, but with absolutely minimal trust requirement and dependency which is nothing like traditional providers.


Anyway, the block size limit should eventually be increased, but not by an algorithm.

rNyMZc18D4yRonb2D9MCjhcHxn8yx9gr31

I know, and I don't know if this can be fixed. The 21M asymptote exists as a gridline and there's a fairly visible gap between it and the graph height in 2033.

I've been pondering the best representation, I think a single digit is too inaccurate (doesn't give a good sense of what the numbers actually mean). 3 digits would have been more clumsy but have the advantage of being exact.

On several occasions I've tried using the popular inflation schedule graph in my presentations, but quickly realized it's not really readable in front of an audience. I therefore designed my alternative version of it:



Key changes:

• The text is much larger.
• The horizontal axis has tick marks only every 4 years. (There is still a grid line every year though.)
• Instead of a vertical tick mark every 3M coins, the 21M range is divided to 8 segments. I think having round numbers is trumped by having the tick marks aligned with the halving points.
• I added dots in every halving point.

I kept the end year of 2033, it's as good as any other.

I'm throwing it here for feedback and for discussion whether it should replace the previous graph wherever it appears.

According to demand I can revert some of the changes or make new ones.

Mathematica code:

I think having to trust specific entities - even if they are multiple and diversified - is against the spirit of Bitcoin.

So that leaves miners and riches, effectively some PoW/PoS combination.

I have my doubts about the consensus model - it looks like a step backwards, what Bitcoin would have been before Bitcoin (and the blockchain) was invented. But assuming it proves itself technically, I'd be open for an alt that uses consensus for synchronization, and PoW for initial distribution (as opposed to Ripple's centralized model).

I thought the OP was talking about "The Rise and Fall of Bitcoin".

He didn't say "ever". He was talking about that particular point in time, and in a specific context.

I regret to inform that there will be no videos from the January 2013 meetup in Google.

Google was able to provide us with a video camera for the event. However, this camera was based on mini-dv cassettes medium which are fairly hard to work with. Additionally, no experienced cameraman was available which means we were left with figuring out how to operate the thing. This led to several mishaps.

The first is that one of the cassettes we tried to use was actually a head cleaning cassette. It has a very short tape and once it finished we replaced it with a good one, so only a few minutes from before the event started was missed.

The two good cassettes had to be physically taken to a photo shop that can convert them to a more accessible format. One shop we tried handing over the cassettes to told us one week later that they couldn't read their format, and didn't provide any other info. Afterwards we tried a different shop, that one told us that they were recorded in NTSC (PAL is more popular here), and that it required a different set of equipment - which they had, but it was a long and costly process (100 NIS per cassette).

When they finally delivered the converted data, I checked it out and noticed to my horror there was no audio. The store confirmed that indeed no audio existed in the recorded tapes, probably because the camera had a microphone that needed to be explicitly turned on but wasn't.

So now we have video footage of the event without any sound, and as fun as it is to watch us making all kinds of hand gestures and wondering what we're saying, I don't think it's of much use. If there's demand I'll consider uploading the videos as they are, but I think it would just cause confusion.

It's a real shame because I think the event was quite successful and I wish we had a way to preserve it and make it available to those who missed it.

To better handle the next events we've now purchased a video camera of our own. It won't be of the same quality, but at least we won't have to scramble calling in favors to get a camera each time, and we can figure it out once and be confident that we'll get it right when it counts.

Confirmed receiving a $20K loan from Vandroiy.

Most of the buying demand is definitely for investment purposes. There's a theory that the Israel boom is (partly) responsible for the BTC rate appreciation, but I don't really believe that.

To be liable you don't need just to own a key, you'd need to be one of the people whose key was lost/stolen.

Also, it's not necessary to have just one big m-of-n group. We could have several groups with wallets of different sizes.

And that's why we're not going to do the former type (in addition to it being much less secure).

If we need to rely on theymos to delete the keys (and that the keys weren't somehow compromised during his generation process), then it's not decentralized.

Issuer risk can be greatly reduced if you use multisig (each party bringing their own key) instead of secret sharing. The issuer's role can then be reduced to compiling the list of keys without each member knowing which key is whose.

You can greatly reduce loss risk, without reducing the collusion factor, by having 4 groups of 4-of-5 rather than 5 groups of 4-of-4, etc.

Also: Your method as described has a lot of variance. (7-way collusion gives access to X coins on average, but on practice it can deviate significantly from average). You can reduce variance by having more wallets, with overlap in custodians (e.g., each member is part of 10 different 4-of-5 groups, total of 40 wallets).

With combinations of wallets of different group sizes, you can control the correspondence between number of colluders and number of coins that can be spent.

Anyway this is getting completely orthogonal to the original purpose of securing funds.

No, it's not.

If I understood you, this begins with the assumption that theymos generates all the keys. If in some way his procedure is compromised, all funds can be stolen regardless of how well everyone else did. If theymos chooses to embezzle (or to just force a dictatorship), he could keep a copy of the keys and steal the funds whenever he wants.

If we do assume the issuer is benevolent, we don't need the complex cryptography - everyone can just tell the issuer what they want and he will do the right thing.


Also: With the method as described, huge risk of loss. If each person has 1% chance of loss, the average loss with "normal" storage is 1%. With your method average loss is 4%.

I'm not sure what exactly you mean but check out

http://blockexplorer.com/
http://blockchain.info/

That I agree with, I guess I misinterpreted the scope of your comment in that specific context.

With m-of-n personal liability is greatly alleviated and thus this can easily be done with no fees. (We could make it so that if someone loses a key but the funds are still safe, he pays a small fine; and if enough keys are lost/stolen to compromise the funds, those who were negligent must together pay up the amount).

If there was no multisig and I had no desire to help the forum and take part in something cool, 3%/year is probably about what I'd charge. Losing those funds would highly correlate with losing my own stash, and in this case I'd be liable and have greatly reduced convenience (not impossibility though) of repayment.

Speaking of which, even with multisig we should be aware of correlated risks. We'd want the different trustees to use as different procedures of generation and storage as possible.

I'm sorry but this is nonsense.

Let's forget about multisig for a second. Storing coins always has a risk of loss and theft. For all offline computer and paper wallets suggestions, physical burglaries exist too and it won't be long until we have Bitcoin-minded burglars. If the offline wallet is further protected with a password, this password is either memorized and can be forgotten (or forced out), or it's written down and can again be stolen. There can also be a sophisticated virus, or simply human error, in the generation process.

(Myself I'm worried that the iso I would download to install Ubuntu on a fresh computer would be corrupted and actually install a compromised OS that intercepts Armory and instead of generating random keys generates keys known to the attacker. I have no idea if that's actually possible, but merely my subjective uncertainty about this possibility means I am taking a risk by holding coins. And those are just the risks I thought of - what about all the risks I didn't?)

The trustee would still be liable for the coins even if they're lost. This means he could easily find himself with a $5K debt he may not be in a good position to repay (and even if he can do it, he still loses $5K). It's very reasonable to require a risk premium for this insurance. Of course, it's also reasonable to take it free of charge as a contribution to the forum.

Why do you think keeping the money with trustees is risky, and keeping it with theymos is not? Utmost respect to theymos, but he could embezzle or lose the funds too and I'm sure there are other people just as reliable.

Let's assume 1%/year chance of a problem. If theymos holds everything there is 1%/year risk of losing all of the forum's money. If it's split with another trusted person, there's 2%/year risk of losing half the money and virtually no risk of losing everything. That's much better because the forum will always have enough funds to do the really important things. It's still better even if some fee has to be paid. In fact the difference is even more pronounced, since the more money concentrated at one place, the greater the incentive to embezzle and the attractiveness for attackers.


Thankfully, multisig solves most of that and allows a solution with minimal risk for everyone.