After Farming


#1

This is a proposal for an incentive model capable of replacing the one based on farming when the time comes. This model adapts this proposal I made regarding blockchain-based cryptocurrencies: http://bitcoinmagazine.com/12296/transaction-rights/.

The proposal is this: When all safecoins have been issued and safecoin has (hopefully) become the dominant form of money in the SAFE network, or even before that:

  • Each node processing transactions will be rewarded with the right to make a volume of transactions corresponding to the volume of transactions it made possible.

  • The nodes with excessive transaction rights can sell those excessive rights to the nodes lacking them. This transaction-right pricing becomes the only form of transaction fees in the network. This way, the fees for those with enough transaction rights (earned by enabling transactions from others) will remain zero.

  • Transaction rights will exceed the volume of transactions enabled by a single node in this way: N*(1+(F/B+R)÷2), where:

    • N represents the volume of transactions of a transaction round: the calculation only takes place when the volume of transactions enabled by the node (excluding its own transactions) reaches N. Then, the result adds to the transaction rights the same node already had.

    • F represents the paid transaction volume in N (F stands for “fee”).

    • R represents the relative price variation of F relative to the previous transaction round, which is (P÷F)÷(Pp÷Fp)−1, where:

    • P represents the absolute price of F (F was the paid transaction volume in N).

    • Fp represents the paid transaction volume in the previous transaction round.

    • Pp represents the absolute price of Fp.

For example, suppose the transaction round has 400 KB of transaction volume. In the current transaction round, 120 KB are paid transactions. This paid transaction volume costed 0.18 coins. The previous transaction round had a paid transaction volume of 96 KB costing 0.24 coins. The relative price variation R is hence (0.18÷120)÷(0.24÷96)−1 = −0.4. Then, the reward for enabling all transactions in the node’s current transaction round will allow it to do a future transaction volume of 400*(1+(120÷400−0.4)÷2) = 380 KB.


#2

This looks like it’s probably a workable idea for blockchain-type cryptos, but I think you need to look closer at how the SAFE network actually works, and specifically how ownership of safecoin is transferred.

Some of the assumptions implicit in your proposal don’t apply to the structure and functioning of this network. For instance, the SAFE network does not have amalgamated transaction processing, i.e., no network-wide blockchain.


#3

Farming never ends. Safecoin will always be recycled from users to farmers in exchange for hosting.


#4

Farming never has to end in general. If it’s divisible then the amount farmed could shrink forever and it would still work. It’s just the recycling method that is favored at this time.


#5

For instance, the SAFE network does not have amalgamated transaction processing, i.e., no network-wide blockchain.

This proposal does not depend on a blockchain. All it requires is:

  1. Defining transaction rounds (to replace transaction blocks).
  2. Measuring the contribution of each node to validating and propagating transactions in the most recent transaction round.
  3. Making the transaction volume allowed to this node depend on that contribution.

#6

Farming never ends. Safecoin will always be recycled from users to farmers in exchange for hosting

According to David Irvine, this is just one of the possibilities:

After bitcoin mining is complete the premise is transaction fees will keep miners up and running. I am not sure this will be the case, however with safecoin there are no transaction fees. The tx fee model is one I did not like. The SAFE community may decide to introduce those, recycle safecoin, increase amount via inflation etc.


#7

This would be adding a very large centralization and complexity of code to what (we hope) will be a very fast and functional system. Just to have a faux block? Can’t imagine why we’d want to do something like this, even if possible. It would be completely disrelated to the actual storage and manipulation of data, and thus be an added drag on the network, to little purpose, IMHO.


#8

It was my fault: I transferred an unnecessary burden from the blockchain-based model when adapting it to safecoin. Each node needs only to accumulate the volume of transactions it made possible and discount from the result of the above calculation when it makes its own transactions. When calculating the proportion of paid transactions and their price variation, the node will again take into account the same transaction volume (including only transactions it made possible). Since each node will do the same, the global result will be the same as if all nodes knew about all transactions in the network. This is an entirely local, decentralized implementation.


#9

Thanks for clarifying. Now that we’ve cleared up that one point, though, you’ve gone completely over my head into realms I know not of. :blush:


#10

I changed the original post according to the result of our discussion.