Truebit Interfaces

[0x3bfc]

Interfaces includes:

• command line interfaces and clients
• Interfaces between layers (incentive, dispute resolution, and off-chain computation layer)

[hswick]

Hi! Thank you for posting an issue. Could you clarify which document you think should include this information?

[0x3bfc]

Thank you for your swift answer. You are really doing great work!

For instance incentive layer has a nice description about the pre-processing and the algorithm, but for developers who are going to integrate their services on top of Truebit, they need more documentation about interfaces.

Also, I'm wondering if there is a documentation for integrating different proof mechanisms instead of interactive merkle proof .... something like ZK proofs. Is there a way to integrate something like this in Turebit Dispute resolution layer ? Moreover I don't understand how DisputeResolution Layer interacts with Ocaml-offchain interpreter ... where I can find such info like this in the code. In short the interfaces between layers are not clear enough to understand.

What If I want to submit my code (WAST/WAT) code or C++/Rust code, I think if this wiki could have a section for some examples to follow, this will be great.

I saw some examples and tutorials which are great like links below:

https://github.com/TrueBitFoundation/wiki/blob/master/docs/WASM/Tutorial.md
https://github.com/TrueBitFoundation/test-node-docker

But they are a little bit distracted, if you can group them in one place this will nice.

[hswick]

Ah thank you for the very in depth feedback!

1. Incentive layer interfaces - yes completely agree. We haven't settled on stable interfaces, but once we do, clear documentation will be crucial.

2. I don't think we have thought about swapping the ZK proofs with merkle proofs yet. The merkle proof is sort of one of the only pieces that definitely belongs in the dispute resolution layer. Could you extrapolate on why you might want a ZK proof instead of a merkle proof?

3. Submitting WASM code - I will make a note about making the tutorial easier to follow. I believe there has been some work done in this regard, so it should be a matter of updating the docs.

[0x3bfc]

sorry for late answer, but I'm really keen to learn more from your tutorials.

Regarding the second point .... I didn't mean exactly swapping zk proof with merkle proof. It was a just a question and I was wondering if it is feasible or not. For example, if we have a solver, verifier, and task giver. The task giver wants to enforce the solver to not reveal any secret about how he does his computation to the verifier and at the same time let the verification game runs as well as the merkle proof. But anyway this was a question and no need for documentation.

Looking forward to seeing your docs :)

[hswick]

No problem! We are working on improving tutorials and usability this sprint, so stay posted.

Ah I see, that would be a nice extra privacy feature. That could be very possible to do in the future with our existing architecture.

Once again, we really appreciate your input!