FRP: Removing the Relay in PBS

[guruvamsi-policharla]

Proposing an FRP on a cryptographic approach to remove the relay in PBS.

[sarahalle]

@guruvamsi-policharla thanks for submitting this proposal! We are reviewing it and will get back to you with any questions or feedback

[nikkisigurdson]

Hello @guruvamsi-policharla. Thank you for submitting this proposal! We are carefully evaluating it, and preparing a detailed response. I will get back to you about this within a few days.

[nikkisigurdson]

Hello, @guruvamsi-policharla. I’m Nikki, I’ll be your main Flashbots point-of-contact on this FRP. There are a few points I would like to see clarified in the proposal, and I have concrete feedback for you below. I am working with people from Product to get information about the preferred starting point for this prototype, and that will be shared soon. Overall, we are excited to work with you and we look forward to seeing a prototype of the scheme get implemented!

After some internal discussion, Flashbots is very interested in specifically removing the relay for BuilderNet/TEE-Boost, so it makes sense to focus the FRP on this application.

The goal of this revisions process is to more clearly delineate the knowns from the unknowns at this time. If any of these suggestions are unclear or impossible to accomplish right now, you can let me know.

The main suggestions for revision are:

1. The 'Background and Problem Statement' section is not very clear. For example, you state that the topic of the FRP is 'Removing the Relay in PBS', but I do not see a proposal to removing the relay for PBS in general; it seems your topic is more specifically removing the relay for BuilderNet/TEE-Boost. Can you revise this section, making it more focused on what background is necessary for understanding the existing role relays play for builders using BuilderNet/TEE-Boost? (Another useful question to answer is "why does the blogpost proposal work for BuilderNet/TEE-Boost but not PBS in general?")

2. The 'Plan and Deliverables' section similarly can be improved by making your plans more concrete and measurable. I have some specific suggestions below that reference the enumerated goals in your proposal:

• Regarding (1), What aspects of the scheme do you plan to prototype, document, and open source in your FRP? The code for the Silent Decryption protocol is already available, so how does that play a role?
• Regarding (2), this information seems more like 'Background and Problem Statement' than 'Plan and Deliverables'. Can you flesh these considerations out more clearly, and highlight exactly what modifications will be needed to consensus, as well as more clearly describe how you plan to document the above changes? (This may be difficult to answer at this time but try your best to collect additional resources. The more concrete, the better.)
• Regarding (3), it is unclear if you are proposing to perform these network experiments as part of the FRP. Can you clarify this point, and if you plan to accomplish these tasks in your FRP (which we encourage), can you give more details about how you will accomplish them?

3. Can you provide a tentative timeline for the various tasks and parts of the project?

If you feel stuck regarding any of these points or have any questions, feel free to ask. As I mentioned above, we’re currently working to get some detailed Product-side feedback on starting points and next steps.

Thanks, and looking forward to working with you.

[guruvamsi-policharla]

Hi @nikkisigurdson,

Thanks for taking the time to review the proposal!
I've updated the FRP with the suggestions.

Here’s a quick summary of the changes:

• I have added more discussion on the role of Relays and how our protocol works for PBS in general. There are two things that need to happen to remove the relay:
  • Privacy of Builders: This is where the new Witness Encryption scheme comes in and works in general for any instantiation of PBS.
  • Safety of Proposers: This step is more specific to the instantiation of PBS. Essentially figuring out the best way to make sure the proposer gets paid.
• #1. Added more details.
• #2. Moved into the main body
• #3. Added clarificaiton

I expect building out the prototype to take up to 2-3 months, as I will not be able to work on this full time. Preliminary benchmarks and Documentation should take another month.

Please let me know if there's anything else that needs to be updated. Thanks!