How, technically speaking, do you plan to implement it in this decentralized system?

I have an idea how to implement this. You raise a few billion dollars reserve and some serious mining and than peg your block-chain to whatever. Market goes higher you sell your coins, market goes lower you buy. Keep doing it until you run out of money. Than peg ends.

This is really the best way I see for implementing it.

Challenge accepted. I'm not a cryptography expert, but I am a software engineer, and I think I can come up with a straw man idea at least for you guys to knock down.

The first thing needed would be a decentralized agreed-upon exchange rate between bitcoins and 1971 US dollars. I don't know how the bitcoin protocol currently does the distributed timestamps, but I'm guessing that something similar could be done for the exchange rate number. For the sake of argument I'm going to assume that having miners encode the exchange rate in their generated blocks would work, requiring a heavy moving average to prevent the exchange rate from moving drastically over a short time. The miners would probably calculate the exchange rate based on the current exchange rates on the biggest bitcoin exchanges, combined with U.S. inflation data since 1971.

Assuming I now have a distributed exchange rate, I proceed to the problem of producing two new bitcoin-backed currencies: 1971coins and hyperbitcoins.

A normal user downloads the shiny new 1971coin/hyperbitcoin client, and wants to convert some of their bitcoins to 1971coins. They see that that the client has a new feature; it says "value stored in: bitcoins change this". They click the "change this" button, and select "1971USD". Their client now says "value stored in: 1971USD change this". Simple.

Behind the scenes, the client sends out a new kind of transaction, paying the client's bitcoins from their bitcoin address (19hMEAaRMbEhfSkeU4GT8mgSuyR4t4M6TH) to their new 1971coin address (45VjRaDX9zpbA8LVnbrCAFzrVzN7ixHNsC). Note that the latter address starts with "4", clearly distinguishing it from all regular bitcoin addresses which start with "1". The bitcoins are not actually transferred - they are held by the network, and can only be retrieved when someone converts 1971coins or hyperbitcoins back to bitcoins.

Eventually, a miner running updated code picks up the transaction and encodes for posterity that this user has surrendered their bitcoins to hold 1971coins instead.

The user then uses those 1971coins like they would normal bitcoins, or stores them in a hole in the ground. No matter where they go, how they are subdivided/combined, they can always be converted back into bitcoins by the inverse transaction, and they will get a different amount of bitcoins from the network at that time (equal to the face value of the 1971coins at the current exchange rate). The bitcoins they receive will be the ones surrendered by other people who converted bitcoins to 1971coins or hyperbitcoins.

Another user converts his bitcoins into hyperbitcoins, using a similar transaction, except he sends them to an address starting with "h" (hFNTpsLm8Mitdd99YC1tHSyN2dVMdQiWYo). His bitcoins are also held by the network. His new hyperbitcoins CANNOT be transferred, spent, etc. They can only be held or converted back to bitcoins. This is because all hyperbitcoins are not of equal value. When hyperbitcoins are created from bitcoins, they get a leverage value based on the current number of 1971coins and hyperbitcoins in circulation. When they are converted back to bitcoins, the miner processing the block looks at the change in the bitcoin exchange rate from the time they were created, and pays out bitcoins according to the leverage of these particular hyperbitcoins times the change in the exchange rate.

In this way, bitcoin speculators (who don't want a stable currency) play the counter-party to people who want a stable currency.

There. That's a shot at how it could work. What I am more interested in, however, is your collective comments on whether this is worth doing.

If a miner tries to put the wrong exchange rate in a block, the block would not be accepted by the rest of the network. The same logic would apply to a miner that tried to put a timestamp on a block that was 5 years in the future. Other nodes would reject that block.

Anyway, I'm trying to get back to my central question. My question assumes that it is possible through some protocol changes. I don't really care what protocol changes.

Assuming it is possible in some way, is this worth doing? Will it massively improve bitcoin utility and value as I imagine it will? Or do you guys think it is just another silly idea that probably can't gain traction?

How much demand would there be for a value-pegged bitcoin-backed currency? How much demand would there be for the leveraged hyperbitcoins absorbing the volatility? Would they balance each other out as I imagine they would?

I really hope not, because anything involving a new blockchain would be pretty worthless, IMHO.

The only way it would be valuable (I think) is to build on the existing block chain, continuing to accept and process standard bitcoin transactions in the usual way while adding new transaction types and rules for the new bitcoin-backed value stores.

Starting separate block-chains threatens the value of everyone's existing bitcoins, so I can't imagine that very many bitcoin people would want to use something new. I'd rather build on the existing bitcoin network and existing coins, and add value to them.

Well, since you've done a lot of coding using the protocol, you would know whether it is possible as well as anybody.

It just drives me crazy that there are so many people who would rather use a stable bitcoin, and there are so many people who want to bet big on bitcoin's success, and there's a very simple arrangement that would allow each of them to give the other what they want! If it only there was a way we could bring those people together in a simple, easy-to-use way, bitcoins would become much more valuable and usable.

Anybody have any other ideas on how to accomplish this? It will probably make bitcoin values go up at least tenfold if we figure out a way to do it.

The post above moves my thoughts out of the context of this thread, so I started a new one here: https://bt.irlbtc.com/view/31032.0