Can crypto-transaction networks function without a built-in currency?

[LudwigW]

Decentralized crypto-transaction networks such as Bitcoin, Ripple, Ethereum and pretty much all others all include their own built-in currencies. As far as I can tell they do this for one or more of the following reasons:

1) as a reward for those who provide their computing power to secure the network
2) as a mechanism for paying the transaction fees that help to minimize spam in the network
3) to provide rewards for the founders (who either pre-mined (part of) the currency or who are among the first to mine the currency)
4) to encourage people to start using the network by rewarding them (if the currency appreciates in value) for buying into the currency that needs to be and can only be used on said network

But are there other ways than by creating a built-in currency in which decentralized crypto-transaction networks can solve these problems? I'm particularly interested in alternative ways of solving the first 2 problems (network security & anti-spam).

[d'aniel]

For (1) you could do merged-mining, as it costs practically nothing extra to miners.  But then you have to be careful to keep interests aligned - the new chain can't be competing with Bitcoin, and should in fact be seen as complementing it in some not insignificant way.

(2) could be addressed using identities that have a cost to create, and can be blacklisted if they are abusive.  E.g. https://en.bitcoin.it/wiki/Identity_protocol_v1, though I'd say just burn the coins in an OP_RETURN <SIN> transaction, since the proposed announce-commit miner sacrifice scheme requires non-standard transactions.

[Foxpup]

You forgot the most obvious reason: a built-in currency is a verifiable thing that can be transacted over the network. If the currency is created and issued by something external to the network, how can individual nodes actually verify that the person spending the currency both a) actually owns it and b) has not previously spent it elsewhere? The only way is to require that the person deposit the currency at a specific institution, but that can't be decentralised.

[LudwigW]

Thanks for your reply.

For (1) you could do merged-mining, as it costs practically nothing extra to miners.  But then you have to be careful to keep interests aligned - the new chain can't be competing with Bitcoin, and should in fact be seen as complementing it in some not insignificant way.

Okay, I think I understand, but with merged-mining the new chain would still have a currency of its own that functions as a reward for miners, no?

(2) could be addressed using identities that have a cost to create, and can be blacklisted if they are abusive.

Yes, good point. Who would blacklist them though in a decentralized network?

[LudwigW]

You forgot the most obvious reason: a built-in currency is a verifiable thing that can be transacted over the network.

Yes. Not sure though that that is a separate reason, more a crucial element that makes a built-in currency able to function as a solution to the problems mentioned in the 4 reasons I gave

If the currency is created and issued by something external to the network, how can individual nodes actually verify that the person spending the currency both a) actually owns it and b) has not previously spent it elsewhere? The only way is to require that the person deposit the currency at a specific institution, but that can't be decentralised.

To be sure, there is a difference between saying that the network requires units of something for transactions and saying that those units function as currency units. I'm not asking whether it is possible to do away with built-in units for transactions altogether, only whether it is possible to do away with built-in units that function as currency units.

If we use the ledger terminology, for the slots in the ledger to function as a currency there need to be limits on their supply. But if they simply function as media for transactions of whatever kind (by representing e.g. external currency units, stocks) then those constraints on their supply are no longer necessary and there can be an unlimited number of them. That way, the spots in the ledger still function as the media that make all sorts of transactions possible (by representing ownership of something external to the ledger) but no longer as currency units. And in that way, individual nodes can still verify that the person selling a spot in a ledger (that represents e.g. a stock, or a dollar) actually owns that spot and has not spent it elsewhere.

The problem then however is that in that case they (especially newly created ones) could no longer be used to reward miners, speculators, founders etc or to prevent spam because they simply aren't valuable in and of themselves anymore. Their only value would be derived from the thing that they represent outside of the network.

So the only way in which they could still function as rewards despite there being an unlimited supply of them is if individual units are tied to something external to the network, and yes, then in many cases issues of trust will re-emerge.

[jaked]

Decentralized crypto-transaction networks such as Bitcoin, Ripple, Ethereum and pretty much all others all include their own built-in currencies. As far as I can tell they do this for one or more of the following reasons:

1) as a reward for those who provide their computing power to secure the network
2) as a mechanism for paying the transaction fees that help to minimize spam in the network
3) to provide rewards for the founders (who either pre-mined (part of) the currency or who are among the first to mine the currency)
4) to encourage people to start using the network by rewarding them (if the currency appreciates in value) for buying into the currency that needs to be and can only be used on said network

But are there other ways than by creating a built-in currency in which decentralized crypto-transaction networks can solve these problems? I'm particularly interested in alternative ways of solving the first 2 problems (network security & anti-spam).

For any network to function, the people running the nodes must be properly incentivized.

Bitcoin has a built-in incentive.

What would your network transact, and what incentive will node-operators have?

[LudwigW]

Decentralized crypto-transaction networks such as Bitcoin, Ripple, Ethereum and pretty much all others all include their own built-in currencies. As far as I can tell they do this for one or more of the following reasons:

1) as a reward for those who provide their computing power to secure the network
2) as a mechanism for paying the transaction fees that help to minimize spam in the network
3) to provide rewards for the founders (who either pre-mined (part of) the currency or who are among the first to mine the currency)
4) to encourage people to start using the network by rewarding them (if the currency appreciates in value) for buying into the currency that needs to be and can only be used on said network

But are there other ways than by creating a built-in currency in which decentralized crypto-transaction networks can solve these problems? I'm particularly interested in alternative ways of solving the first 2 problems (network security & anti-spam).

For any network to function, the people running the nodes must be properly incentivized.

Bitcoin has a built-in incentive.

I know and my question is whether there are other ways to incentivize node-operators than by using a built-in currency

What would your network transact, and what incentive will node-operators have?

the idea is that the network would function as a ledger the same way Bitcoin does but while bitcoin has a limited number of units that can be used for transactions this network would have an unlimited number of units. That makes those units useless as a currency and so valueless in and of themselves. Instead they only come to have value bwhen they are used to represent   something external to the network (e.g. a stock, a dollar, a bitcoin), and so transacting in the unit is transacting in the thing that it represents outside of the network. (similar to how e.g. colored coins work)

Because there would be an infinite supply of units (but of course only a limited number of units that each represent e.g. a stock in company X) they are not currency units and value-less in and of themselves, so they can not be used to reward e.g. node-operators. So the question then is whether there would be other ways to reward node-operators.

[Foxpup]

You haven't answered my question. How do nodes verify that a person making a transaction actually owns the external unit? And how do you prevent that person from spending that external unit elsewhere outside of the network? Since it's external, and the transaction only happens on the network, the unit itself hasn't actually changed hands. The original owner still possesses it, and the new owner has no way of physically obtaining it. Explain how this problem can be solved without requiring a central depository.

[LudwigW]

You haven't answered my question. How do nodes verify that a person making a transaction actually owns the external unit?

they can't (at least not by only using only tools available within the network), and that is where trust issues re-appear.

And how do you prevent that person from spending that external unit elsewhere outside of the network?

same story. but this holds for all metacoins or alternative networks (such as Ripple)

Since it's external, and the transaction only happens on the network, the unit itself hasn't actually changed hands. The original owner still possesses it, and the new owner has no way of physically obtaining it. Explain how this problem can be solved without requiring a central depository.

well, you don't need a central depository, but you do need to trust e.g. the company that issues stocks (when each stock is represented by one unit within the network), the company that stores your gold etc

at least the public ledger provides transparency and will limit people's and businesses' abilities to deceive the people they transact with.

So i guess there is a fifth reason for including a built-in currency, namely that as long as it is only used as a currency (and not as a vehicle for a metacoin) it eliminates the trust problem which is something that doesn't work for an external currency represented in the network by a unit

[LudwigW]

there is also a sixth reason: if the network is going to function as an exchange as well, then a built-in currency can function as the bridge currency (Ripple mentions this reason)

[Sukrim]

You haven't answered my question. How do nodes verify that a person making a transaction actually owns the external unit? And how do you prevent that person from spending that external unit elsewhere outside of the network? Since it's external, and the transaction only happens on the network, the unit itself hasn't actually changed hands. The original owner still possesses it, and the new owner has no way of physically obtaining it. Explain how this problem can be solved without requiring a central depository.

Well, as you even say so: stuff that is EXTERNAL to the network cannot be controlled BY the network of course.

The central depository then could be the network itself for example, as long as money is viewed as something (ultimately) physical that cannot be created by anyone within some limits though, this is going to be very difficult. If the thing transferred is pieces of data, computation or something like that, you might have better chances.

[LudwigW]

I guess it would be possible for the network or applications built on top of it to transfer control over or access to some types of things external to the network (by e.g. transferring the digital key needed to open a safe, operate a car etc)

[d'aniel]

Thanks for your reply.

For (1) you could do merged-mining, as it costs practically nothing extra to miners.  But then you have to be careful to keep interests aligned - the new chain can't be competing with Bitcoin, and should in fact be seen as complementing it in some not insignificant way.

Okay, I think I understand, but with merged-mining the new chain would still have a currency of its own that functions as a reward for miners, no?

No.  Remember: it costs practically nothing extra to merged mine the new chain, so the incentive to mine doesn't have to be very large - complementing (and not competing with) Bitcoin in some not insignificant way could suffice.

(2) could be addressed using identities that have a cost to create, and can be blacklisted if they are abusive.

Yes, good point. Who would blacklist them though in a decentralized network?

If one of your peers is being abusive, you would blacklist it.  Notice though that a single identity gives you the ability to abuse every node once - ideally you want it so that abusing one node gets you on every node's blacklist.  Distributing cryptographic proofs of bad behaviour might a way to achieve this.