how does the loan work?

[owocki]

i’m not sure i understand how the loan works. why a loan? who manages it? whats the flow of funds over time?

could use a state diagram of that part

[MadeofTin]

The paying back of the loan is completely automated by a smart contract. The block rewards are sent (per block) to a smart contract that deposits the funds into the lending accounts. The Funding organizations are able to withdraw from their account at anytime. Here is a state Diagram I hope is more clear. I will add it to the proposal after settling on the design.

[owocki]

questions:

1. as a 'funding organization' how are things handled if this EIP is not accepted by the miners?
2. do you have a copy of hte smart contract for "lock/unlock"ing the funds. what is the cost / timeline of developing such smart contract?
3. how is the smart contract "unocked" by Istanbul? an irregular state change?
4. how did we arrive at the 124k ETH number? does it make sense to start with a smaller amount, say, 1k ETH?
5. who controls the dxDAO and how is that governed?

[MadeofTin]

1. as a 'funding organization' how are things handled if this EIP is not accepted by the miners?

Could a miner nit-pick this change out and block it somehow? Collecting signals from miners before it is accepted will help at least get the sentiment, but I didn't realize this this would be possible to censor. Afaik from conversations with the DECRED guys it isn't possible for a miner to stop them. Of course not accepting the Istanbul Hardfork is a different story.

The safe option then would be to unlock the funds through an irregular state change that is also accepted at the same time the block rewards are accepted. So they could accept both or reject both, but either way the funds can be recovered.

2. do you have a copy of hte smart contract for "lock/unlock"ing the funds. what is the cost / timeline of developing such smart contract?

There are some implementations floating around. The simplest way would be to collect them into a multi-sig, which that has its own human risks, but I think it is worth exploring this as a mechanism to keep the funds safe. https://github.com/chatch/hashed-timelock-contract-ethereum

3. how is the smart contract "unocked" by Istanbul? an irregular state change?

Either an irregular state change, or it can be after a certain number of blocks.

4. how did we arrive at the 124k ETH number? does it make sense to start with a smaller amount, say, 1k ETH?

We only have an opportunity for this every hard fork, so in essence we only really have one chance to get this right. The risk of underquoting is higher to me than over quoting. I would rather make sure we have enough to finish 1.X and then burn extra funds, then not have enough and fail at raising more because we failed to raise enough funds the first time.

The number is in line with conversations I have had with funds needed for a project with the scope of Eth1.X. It also matches well with the EF spending projections. 30 million this year extended to 18 months would be 45 Million. So 20 million into development is not an unrealistic number to raise.

There needs to be some room allowed for market movements and I would also recommend for the Devs to keep 60% in DAI and 40% in Eth as the project progresses. This can be rebalanced monthly and perhaps the left over Eth in the end can be burned. This may also be worth explicitly putting in the proposal.

So in essence raising 20 million in Eth, isn't really raising 20 million. It is raising +/- regular 25% volatile swings over a period of time with a consistent drain on those funds rain or shine. Better safe then sorry.

The other option would be to space out collecting funds over multiple Hardforks. 1K or even 10K Eth would only fund Eth1.X partially. It may be easier to get a smaller amount first, but it is also to me a commitment to underfund them. Eth1.X doesn't need to be funded in 1 to 2 years it needs funding for the next 1-2 years.

The primary reasons not to raise the full amount for Eth1.X are three:

1. Inflation on the network
2. Security of the fund holding mechanism
3. Funds being mis-spent

1. At 0.1% inflation 124K ETH is a small amount in relation to the network.
2. While we can look at other mechanism in the short term, there exists today mechanisms that have already been battle tested we can fall back on.
3. Requiring Transparent use of the funds keeps the Devs accountable to the community.

All of these are solvable problems to me. Yes the last one is still in progress of how it will be solved, but it still is solvable.

5. who controls the dxDAO and how is that governed?

The governance would be through the holographic consensus mechanisms has already been tested by daoStack. The Eth1.X core devs themselves would be the managing members.

The scope of the DAO

• Proposals for funding would be made and accepted
• Proposals for membership would be made and accepted.

The other option would be a multisig with the leads of working groups.

[owocki]

Of course not accepting the Istanbul Hardfork is a different story.

This is the condition I’m woried about.

The number is in line with conversations I have had with funds needed for a project with the scope of Eth1.X

Ah interesting; the google spreadsheet I put together doesn’t list anywhere near this number. But then again, not every working group filled out budget. And it was only scoped to be Istanbul + 1 Hard Fork.

[MadeofTin]

Of course not accepting the Istanbul Hardfork is a different story.

This is the condition I’m worried about.

This is the major justification of the interest. In between the deposit and the start of the fork there is a time paying back these funds will be at risk.

• If it doesn't go through the funds that haven't been spent can be returned at a small loss. The time between the fork and the deposit of the funds is only a few months.
• If it goes through then the orgs receive 10% on their deposit.

We could even formalize that no more than 5% or 10% of the fund can be spent before the fork is accepted by the miners.

The orgs have the chance of no greater than a 5% loss, but the opportunity for a 10% gain.

The number is in line with conversations I have had with funds needed for a project with the scope of Eth1.X

Ah interesting; the google spreadsheet I put together doesn’t list anywhere near this number. But then again, not every working group filled out budget. And it was only scoped to be Istanbul + 1 Hard Fork.

I would appreciate more feedback on this. It is unlikely we even have all of the working groups decided, so this is really difficult to get a good answer for. Some new groups will spawn and some will end and then be replaced by a different ones. My calculations where based on a likely coordination limit of the Eth1.x devs plus a wide margin of error to account for price fluctuations and some accounting for in-person meetings and travel.

[MadeofTin]

Also, after these comments I have also decided to really simplify the model. A simple smart contract that distributes the incoming block reward is sufficient

How/where the funds are collected I am not as settled. Honestly, It would be great if the EF could act as an intermediate custodian.

[MadeofTin]

I simplified it even further and wrote a spec where it is coded into the clients itself. A little bit of a less "elegant" solution, but arguably more secure and and testing will be able to catch if any of the values are wrong.

[lrettig]

Couple of quick questions:

• Is this a "tipping point" fundraising mechanism (a la Kickstarter)? I.e., if the target funds aren't raised, then no funds would be disbursed. Or else can the funds be partially raised? Is this written in code or handled manually? (Depending on the window for collecting funds, there may not be enough time to finish fundraising before the Istanbul code freeze.)
• You mention a "DxDAO" in the previous image - is that still the plan?

[MadeofTin]

The DXDao is now a multisig, any dao would need to be part right part of a longer term plan.

If We don’t raise the funds in time, At the time of an instanbul code freeze we can also freeze the loan terms. I don’t think we will need a kickstarter mechansim. We could do a target of 50% where won’t move forward if We don’t pass that threshold and have a max amount raised.

[ethernian]

Usually, it is an investor, who puts his money at risk and makes management decisions. The failed project is his personal loss. "Funding Organization" gets his profit regardless of whether a project will succeed or not. What will they get their money for?

Who has losses if some project will not deliver in time and quality?
Who has incentives and responsibility to make the best project from DxDAO's money?