Topic: Scalability multiple currencies (Read 9139 times)

That is the point, they *are* the same.   I was making an analogy as a means to show how the system could be managed in parallel with only periodic "synchronization" points.  As it is now they whole system must be synchronized and thus all transactions broadcast to all nodes. 

I don't see the point of them.  Just sounds like bitcoins with a different name, but no actual difference.

You are not pegging two different things (gold/silver) you are simply coming up with two representations for the *SAME THING*.   The *only difference* is which block list is currently tracking the transactions.

So each block list would need to verify transactions to/from the other block list.   Thus you would have some distributed "bridge" nodes that are actively listening to the block lists on *both networks*. 

So I think the problem is with the subject "multiple currencies".   I am really talking about one currency with a split block list and some automated P2P verification of all transactions from one block list to another.   

The problem is that smaller networks are easier to overwhelm with computer power than one big ever growing block list.    It will reach a point where only "big companies" can afford to connect to the block list and all transactions to smaller nodes would have to go through them.   At this point the entire system can be controlled by a government taking out a couple dozen targets. 

I totally understand that you cannot FIX PRICES whether that is an interest rate or a ratio between any two goods. 

I don't see that it couldn't be done. After all China pegged the currency for a billion plus people against the dollar. If you have a central point of control (conversion point) anything is possible.

That is not an argument for it being a good idea. Should Greece be pegged to Germany? Maybe...?

I really don't see how a 1 to 1 conversion is even sustainable between two "currencies" even if somehow they were linked.  The only way it could even be remotely stable is if they were serving two different populations, and even that would be unsustainable.

Here is the issue:  Two separate currencies would be presumably competing for the same "market share" in terms of their use and application.  The slightest difference between the two currencies would then be exploited by somebody or anybody to drive one out of the marketplace and have the other currency then become the dominant currency wholly supplanting the other one.  It could go back and forth for awhile, just like the 16:1 silver to gold ratio that the U.S. government maintained for some time until it too became unsustainable.

The gold analogy is also inappropriate for bitcoins as the only value that is proven with newly "minted" (aka "generated") bitcoins is a "proof of work".  It is important to note that an individual bitcoin doesn't really represent CPU bandwidth or anything else, but merely pure and blind luck.  The coin generation scheme is only a coin distribution model that keeps the distribution from a central authority rather than being kept in the control of a single individual for  political favors.  If there is a "backing" to bitcoins, that would be something called "luck".  Some people simply are luckier than others.  You can tweak that luck with more CPU processing power or coming up with a better algorithm optimized for generating the bitcoins, but it really is simply luck alone that is backing up the money.  There is no gold.

I don't know how you can quantify luck in any other way, and certainly that is non-transferable other than as spelled out in the current rules of the Bitcoins network.  Attempting to reconcile any attempt to maintain a one to one conversion with another similar currency system would simply turn the nodes on that other network into nodes on the current network.

Ah, I see what you are going for now. I don't really know what use case you are thinking about.

Say you duplicate the block list entirely, then split the user base across two non-interconnected systems forking the block list from that point on. Interestedly, all the transactions remain consistent enough to be concatenated. They couldn't actually be concatenated because of the proof of work daisy chain, but you could certainly rerun each fork into the other with no inconsistencies or double spent coins.

The restriction is that each private key/bitcoin address would have to trade on one system or the other but not both. If you merged the transaction list periodically you could allow users to cross over at that point.

Have no idea if this solves any problem, much less the one you were interested in. But I found it interesting to realize that you don't have to destroy the coins. You have to destroy the owners! :-)

To implement this would likely require a breaking change *or* a trusted 3rd party that will hold a coin in a wallet and issue a new coin in the other system.   However, I do not see why this "trusted 3rd party" could not be replaced with a distributed transaction auditing system like is used in bitcoin all ready.   In fact it should be possible (if you have both block lists) to follow the transaction history back and forth. 

1 to 1 conversion is possible if there is a means of "melting a gold coin and re-minting it".   I use gold to represent the "value" and the mint/design pattern to represent the form.

The goal of this exercise is how can you divide the block list in half for an extended period of time and be 100% positive that all transactions within either block list are valid and ultimately portable back and forth between the two systems.   

You could issue two currencies that are only convertible in a "market of free exchange", but that would greatly increase the cost of conversion.

The problem with "floating exchange rate" is that now people need to mentally track and manage the risk.

The second "parallel" system would NOT MINT NEW VALUE FROM NOTHING.   

I do understand the alternate coins with a floating conversion rate. You list some great examples.

I don't understand the alternate coins with a one-to-one mapping to bitcoins. I may be missing something.

For something like this to be effective, it would have to be something that has an arbitrary floating exchange rate, not something with a fixed rate.  Otherwise all you are talking about is a re-denomination of bitcoins and that is something which has already been talked about to a rather large degree.  In other words, having a "Millibitcoin" that can be exchanged at the rate of 1000 mBC per BTC.  That really isn't a network design but rather a user interface issue and no new coins would really be created.


I have talked about an MMORPG situation here:

https://www.bitcoin.org/smf/index.php?topic=353.0

This would require alternate currencies as a part of the game play, and those currencies would be capable of being exchanged with Bitcoins at a floating rate.  The purpose of creating these "currencies" is mainly for creating a complex virtual world that is founded upon some sort of sound economic principles.  Note that this is something that is seen as a completely separate application and not something that is intended to become a part of the mainstream Bitcoins client interface.

Perhaps if there were some disgruntled Bitcoins users that wanted to cause a fork of the network with different monetary policies there could be some sort of completely different currency that could be created with similar principles to Bitcoins, or some situation where there would be substantial breaks in networks "forcing" two different electronic currencies such as what I talked about here with this "Bitcoins on Mars" thread:

https://www.bitcoin.org/smf/index.php?topic=506.0

While this thread sort of devolved into silliness, I do anticipate that there will be people on Mars before the final new coin generation block is created.  In other words, within a hundred years it may become a legitimate issue.  Perhaps if China set up the "Great Firewall" to start blocking external (to the country) Bitcoins data blocks that this could become a country that would have an isolated network of something like Bitcoins that would have to be exchanged at some sort of exchange rate to "foreign" currencies.  It does take a stretch of the imagination to see when this would apply to the current situation.

I think you understand completely. Since there is no way to destroy bitcoins they would need to be held in a trusted escrow account. That is the only way to assure there will be bitcoins at the correct market rate to exchange back.

I'm missing the point of the exercise entirely. How does this additional complexity add a new feature to the existing implementation?

I'm very much new at this and still struggling to understand the full technical scope of BTC, leave alone the social and financial implication, but nonetheless this multiple linked currencies idea seems very interesting in making the system somewhat more resilient to government intervention, as multiple parallel networks could exist.

But the one thing I don't understand is how the 'destruction' of one currency work on the exchange for another... wouldn't that imply a central 'wallet' holding those blocks? Isn't this kind of hard to achieve using the distributed p2p approach?

I know an 'institution' trusted by all involved parties would probably need to exist, but that institution would, in effect, hold *all the negative balance* of transactions made between currencies, only as positive balance... or am I missing the point completely?

You can have a fixed exchange system because the two systems would talk to each other to ensure that any new coin minted in one system results in a coin being destroyed in the other.  Because the transaction is always reversible for free at any time on demand by the user a user would not care which design is on the gold coin they were paid in.

You need not track both systems if you are only doing exchanges in the one system.

The *key* is convertibility between the systems.

If there was no "demand" for bytecoins then every bytecoin would be melted down and reformed as a bitcoin.

However, if the system was set up such that re-minting was always possible at any time then the coins would have equal value.   

Thus each country could have their own "gold coins" but even though they have different names and denominations they all ultimately have equal purchasing power. 

If for some reason people just perfer the look of the US gold coin vs the canadian one and melting/recasting was free then the canadian coin would go out of style. 

So you have your two independent block lists representing bit and byte coins.  Some nodes would have to track both block lists.  A transaction would be entered at the same time in both lists to destroy one coin and create another.   


Put yet another way.  If I take one gold coin out and divide it up into 1000 pieces called gold bbs.  I only need to ensure that the bbs are not "double spent" and I could care less about the other gold coins in circulation.  At any time I could convert my gold bb back into .001 gold coins.  It is only when I decide to "hop boundaries" that I need to ensure that there is no "double conversion". 

I am not sure if you can have a fixed exchange system. If the demand is sufficiently different, than this warrant a floating market exchange rate.

How does your solution solve the problem in the OP?

It seems to be that instead of just tracking all bitcoin transactions I now have to track all bytecoin transactions as well. Which isn't that difficult if I'm determined.

I was implying that the two systems are linked with a built in fixed exchange.

At any time you can convert your bit coin into a byte coin and then back to a bit coin on a 1 to 1 ratio.    In other words, the cost to mint a byte coin is one bit coin.  Every time a byte coin is minted a bit coin is destroyed.   Likewise, the cost to mint a bitcoin is one bytecoin and the bytecoin is destroyed.    Think of it as melting down a US gold eagle and reforming it into a canadian gold coin.  Both coins are 1oz of gold. 

Now the byte coins can go on trading on their own block list without having to check in with the bit coin users.  Value can be transferred between the two systems when necessary, but this is a low level synchronization point.


What I described above is very different from what you described which would be introducing a new "silver" currency to compete with the "gold" currency.  In such a case you would have a floating exchange rate based upon market value of the two systems.   

If Bitcoin takes off it will be interesting to see what similar systems are created. Approach the limit faster? slower? More frequent block generation? Whatever it is will have to overcome the head start Bitcoin has. It would have the advantage though of people being able to buy them with already established BTC instead of PayPal or such.

However an alternative is implemented tough there is no reason to think it will have the same value as BTC. In fact if someone started an identical system it would probably have less value forever.

The difference between the two coin supply and demands entails another market exchange currency, which is an another playground for me.

Bit coin allows us to track all transactions for a particular "coin" type.   

Now imagine a second "coin type" with its own set of nodes only this coin type is "convertible" to and from actual BTC with a special transaction among the two trees.   Now  you have two parallel sets of coins that have the same value, but require an extra transaction to port from one chain to another.   You cannot mix and match coins of different types in the same transaction, but you can at any time convert from one coin type to another.