to put a proposal to a vote in a snapshot, you need to have 500,000 ARB. Do you have them?
Quite a lot of participants can do this for you if you ask them. This is not the first time this has happened. Only about 50 delegates can do this
If at least one agrees with your point of view, there will be a vote
No one needs it. This is the whole power of the DAO. Management is a myth, a fairy tale. The DAO is managed entirely by the Arbitrum Foundation. The staking offer from PlutusDAO, which was approved by the DAO but rejected by PlutusDao themselves (God, how funny this is), reflects this well.
I think this dispute will lead nowhere.
But you answered your own question if you think that no one needs it.
Based on feedback, research, and further development, we have made some updates to the BoLD AIP. Key highlights include:
-
Reducing the challenge levels from 4 to 3 and reducing the challenge bond sizes by ~43% without compromising on security! This means that previously, it would cost 1000+100+10 ETH to fight off 1 challenge, but now it will only cost 555+79 ETH! These new numbers were carefully selected using the formulas and specifications in the whitepaper - optimizing for security (resource exhaustion ratio) and the worse-case scenario of elevated, sustained L1 gas costs during an attack. The top level assertion bond of 3600 ETH remains unchanged because it acts as a deterrent to open a challenge and spam the protocol to begin with.
-
A “defender’s bounty” of 1% of confiscated bonds has been added to incentivize honest parties to defend Arbitrum One in the event of a challenge. This 1% would only go to defenders who put up bonds to defeat malicious actors during a challenge. Since confiscated funds will be sent to the ArbitrumDAO treasury, a DAO vote would be required to pay out this reward though because confiscated funds (by default) go to the DAO.
-
An upper bound on L1 gas cost reimbursements to defenders has been added and is proposed to be 100% of confiscated bonds during a challenge. Honest parties should always be made whole when defending Arbitrum.
-
Challenge bonds are now automatically refunded in-protocol once a sub-challenge assertion has been confirmed by the protocol!
6 Likes
@404DAO’s understanding is that Arbitrum DAO will play a role in deciding what to do with confiscated funds from a malicious actor. We would like to get the conversation started (if one hasn’t already begun) around how best to ensure that the DAO makes an educated decision.
The proposal authors present a few possible options that the DAO can choose from when deciding how to use the confiscated funds:
- Using the funds to refund L1 gas costs to honest parties,
- Rewarding or reimbursing the honest parties with some, or all, of the confiscated funds,
- Burning some, or all, of the confiscated funds, or
- Keep some, or all, of the confiscated funds within the ArbitrumDAO Treasury
We are not proposing any options yet (e.g. create a new committee, use an existing committee, or something else) but would like to plant the seed for a future conversation.
3 Likes
This is one of the more exciting developments occurring in Ethereum L2 land - can’t wait to see it go live. Stage 2, here we come!
I read the full proposal, all the comments, and the FAQ linked above (still need to watch the governance call recordings). However, my one concern comes down to paying out proposers in ETH - targeting the ETH staking yield. I understand the idea is to compensate proposers for the bond / opportunity cost associated with that (which is of course the eth staking yield), but I do not love the idea of having service fees denominated in a currency that we cannot control and have a small amount of in our treasury.
This is the classic emerging markets problem where nation states have a sovereign fiat currency that devalues versus their USD denominated debt obligations, which requires more native currency creation to pay USD liabilities and ultimately devalues the native currency further.
In this example, ARB is the EM’s sovereign fiat currency and WETH is the USD debt obligation.
I do not have a good solution. On the one hand, we want a diverse set of proposers further down the line… but we cannot do that under the proposed structure without scaling ETH costs (with ETH revenues ironically on the decline). On the other hand, if the service fee is paid in ARB, and the USD (or ETH) value of that ARB decreases, the SECURITY of the network decreases in tandem. The latter is clearly the worse option between the two.
I am all for passing this proposal as is, but the DAO must proactively consider this to ensure the long-term sustainability of permissionless validation via BOLD.
1 Like
We voiced an opinion during the call but for the record, we would like to bring up a point that the DAO needs to consider in the future; with the current specification, we don’t necessarily incentivize honest operators running BOLD validators even though it’s quite a large commitment from them. “Defender’s bounty” is an interesting idea, but we aren’t so certain that it’s enough for serious operators to join the operations and achieve some level of diversity.
There are other points that brought up from the delegates and we should continue to discuss them after the release of BOLD.
How does BOLD ensure fair participation and prevent monopolization by well-resourced entities given the high bond requirements for validators?
Generally speaking, we are in support of the network upgrade and have no extra technical points to chime in here that haven’t been discussed or laid out.
When it comes to what Rika messaged about confiscated funds from malicious actor(s), our opinion is that these funds should best be used to refund L1 gas costs to honest parties. Ensuring this isn’t a cost to these parties helping run the network is top priority. Hence the 400 ETH refund for L1 gas costs for proposers is something we are voting a strong yes on. The other 500 ETH on the other hand for proposers who are actively participating we are a little more comprehensive on. We would like to see a breakdown of the funds and which parties are getting what.
As for the third vote that is up on snapshot, we think it makes sense to adopt the foundation as the first active proposer for Arbitrum One. We only need one honest proposer as any one time and it makes sense to have this crucial responsibility be on the foundation. As for the funding, once again, it would be nice to get a breakdown of all the funding. A reason why we are leaning rather positively on this is because of the refunds clause within 30 days for extra funds.
This proposal has now been raised to Snapshot for a Temperature Check - Snapshot
1 Like
It’s great to see such vibrant back and forth around what is one of the most crucial updates to the network since inception!
We just wanted to point out that, for those looking for a high-level overview of dispute resolution mechanisms in the ORU landscape, we just published a comparative analysis as part of the ARDC.
BoLD of course is included but the report also looks at Fuel’s hybrid proving, Optimism’s FDG, and Cartesi’s DAVE.
6 Likes
BOLD allows for permissionless validation, removing the reliance on a permissioned set of validators and enhancing decentralization. It also fixes the challenge period, preventing delay attacks and ensuring timely dispute resolution. These improvements benefit all Arbitrum users, node operators, dApps, and bridges.
1 Like
I strongly support the AIP to bring the BoLD upgrade to Arbitrum One and Nova. This new dispute resolution protocol will significantly enhance the security and decentralization of the Arbitrum chains by enabling permissionless validation and mitigating the risk of delay attacks. The thorough testing and audits, along with the clear economic incentives for honest participation, ensure that BoLD is a robust and reliable solution. By adopting BoLD, we can take a major step towards solidifying Arbitrum’s position as a Stage 2 Ethereum rollup, benefiting all users and contributors to the ecosystem.
2 Likes
Voting in favour of this temp check related to bringing BoLD to arbi one and nova, relying to also the technical opinion of Delphi here. No point in gating the upgrade of the protocol in general.
I want to thank other delegates for sharing more technical details on the design choices behind this system (re: @RikaGoldberg mentioning usage of confiscated funds, @swmartin about a potential divergence of viability if there is a big divergence in future of eth price compared to arb, and @PGov for the request of breakdown of funds distribution). Most if not all of this are takes likely outside the comprehensive skillset of a lot of users, and reading them helps understanding potential pitfalls for the future.
EDIT: realizing this is the topic to comment on also the other 2 proposals,
Voting in favour of funding the first BoLD validator and making the foundation run it. We need at least one honest validator, and to be honest we can’t find a better aligned stakeholder than the foundation to do it. Having them running it, with even the caveat as per the proposal that we will be able to claw back funds, is the safest scenario i can think about for now for our ecosystem.
Finally, voting in favour of funding the operational costs for the BoLD validator run by the foundation. The proposal makes imho a good breakdown on the cost related (3% of the 3600 eth for 3 years, so 432 eth, plus the buffer of 68, + 400 eth for gas costs to post and challenge + 500 eth for service fees). Up the point a simple cow is able to understand it. Also foundation won’t take service fees from this.
As a final note, we likely want to keep a tab on all of these numbers and more in relation to the discussion we are having about using the sequencer revenue. While I don’t think is written anywhere, my educated guess is that we will take the eth from there indeed (if not and I missed something feel free to correct me). We finally got an excuse to use them i guess.
4 Likes
There will be an ‘Office Hours: BoLD AIP Temperature Checks’ governance call later today (as listed in the governance calendar), to cover the 3 on-going BOLD-related temperature checks. See details below:
Office Hours: BoLD AIP Temperature Checks
Monday, June 10 · 3:00 – 3:45pm
Time zone: UTC
Video call link: https://meet.google.com/ktv-dqwg-nne
Preventing monopolization is not the goal of BoLD’s design. As mentioned in the BoLD AIP, the documentation about BoLD, and the BoLD whitepaper, a BoLD-enabled chain only requires 1 active, honest participant at any point in time to advance and secure the chain. Furthermore, there are trustless bonding pools that anyone (you included) can use to deploy using any amount of ETH alongside an assertion you believe to be correct. All in all, participation in the BoLD protocol is permissionless (anyone can do it since Arbitrum node software is public) and we believe trustless bonding pools are a great way to lower the resource requirements for people to participate.
We believe that the proposal as-is can be voted on and, if passed, can be safely deployed to Arbitrum One and Nova without having a polished, finalized economic model. As mentioned in the BoLD AIP, the documentation about BoLD, and the BoLD whitepaper, a BoLD-enabled chain only requires 1 active, honest participant at any point in time to advance and secure the chain. There is no technical need for multiple proposers and anyone can permissionlessly challenge a proposer’s assertions using their own funds or via a trustless bonding pool. What is the concern that you’re interested in resolving via “achieving some level of diversity” for the proposer role?
2 Likes
Arbitrum One and Nova, as well as the implementation of the proposed BoLD upgrade, relies on Ethereum for security. Using ETH for service fees aligns the incentives for actors who both secure and use Arbitrum One and Nova since those same actors are implicitly also relying on the censorship-resistant and decentralized properties of Ethereum. Check out the BoLD FAQ for more information on why ETH is the proposed bonding token of choice for BoLD.
2 Likes
As proposed in the BoLD AIP, the DAO will prioritize making honest defenders whole (i.e. reimbursed for their gas costs for defending Arbitrum One) and paying out the defender’s bounty (1% of confiscated bonds). Any remaining funds can be treated as net income to the ArbitrumDAO treasury. We’re looking forward to seeing where those discussions go!
3 Likes