🚧 This EIP had no activity for at least 6 months.

# EIP-3267: Giving Ethereum fees to Future Salaries Source

Author Victor Porton, Victor Porton https://ethereum-magicians.org/t/discussion-of-eip-3267/5343 Stagnant Standards Track Core 2021-02-13

## Simple Summary

Transfer a part of Ethereum transfer/mining fees to Future Salaries contract

## Abstract

Transfer a part (exact fractions - TBD) of mining/transfer fees + (probably: TBD) some minted ETH to the DonateETH contract configured to transfer to SalaryWithDAO contract.

## Motivation

This proposal solves two problems at once:

1. It provides a big amount of “money” to common good producers. That obviously personally benefits common good producers, allowing them to live better human lives, it increases peoples’ and organizations’ both abilities and incentives to produce common goods. That benefits the humanity as a whole and the Ethereum ecosystem in particular. See more in the discussion why it’s crucial.

2. This would effectively decrease circulating ETH supply. The necessity to decrease the (circulating) ETH supply (by locking ETH in Future Salaries system for a long time) is a well-known important thing to be done.

Paradoxically, it will directly benefit miners/validators, see the discussion.

## Specification

(TBD)

SalaryWithDAO = TBD (address)

DefaultDAOInterface = TBD (address)

MintPerPeriod = TBD (uint256)

TransferFraction = TBD (0..1)

MineFraction = TBD (0..1)

The contracts source

Prior to FORK_BLOCK_NUMBER, SalaryWithDAO and DefaultDAOInterface contracts will be deployed to the network and exist at the above specified addresses.

Change the Ethereum clients to transfer at every ETH transfer and every ETH mine a fixed fraction TransferFraction of the transferred ETH and MineFraction of the mined ETH to a fixed account (decide the account number, it can be for example 0x00000000000000000000000000000000000000001 or even 0x00000000000000000000000000000000000000000 or a random account).

Change the Ethereum clients to mint MintPerPeriod ETH to the contract DonateETH every some time (e.g. the first transaction of the first block every UTC day - TBD how often).

Change the Ethereum clients to every some time (e.g. the second transaction of the first block every UTC day - TBD how often) transfer the entire ETH from this account to the contract DonateETH.

Because this EIP solves a similar problem, cancel any other EIPs that burn ETH (except gas fees) during transfers or mining. (TBD: We should transfer more ETH in this EIP than we burned accordingly older accepted EIPs, because this EIP has the additional advantages of: 1. funding common goods; 2. better aligning values of ETH and values of tokens).

## Rationale

The Future Salaries is the only known system of distributing significant funds to common good producers. (Quadratic funding aimed to do a similar thing, but in practice as we see on GitCoin it favors a few developers, ignores project of highly advanced scientific research that is hard to explain to an average developer, and encourages colluding, and it just highly random due to small number of donors. Also quadratic funding simply does not gather enough funds to cover common good needs). So this EIP is the only known way to recover the economy.

The economical model of Future Salaries is described in this research article preprint.

Funding multiple oracles with different finish time would alleviate the future trouble that the circulating ETH (or other tokens) supply would suddenly increase when the oracle finishes. It would effectively exclude some ETH from the circulation forever.

## Backwards Compatibility

Because transferring to the aforementioned account is neither mining nor a transaction, we get a new kinds of ETH transfers, so there may be some (expected moderate impact) troubles with applications that have made assumptions about ETH transfers all occurring either as miner payments or transactions.

## Security Considerations

The security considerations are:

• The DAO that controls account restoration may switch to a non-effective or biased way of voting (for example to being controlled by one human) thus distributing funds unfairly. This problem could be solved by a future fork of Ethereum that would “confiscate” control from the DAO.

See more in the discussion.

Copyright and related rights waived via CC0.