[Core Protocol] Prove and finalize pre-bedrock withdrawal in 2025

[lhcadmins]

Did you check the documentation and specs?

• Yes

Was there anything unclear or missing?

I'm trying to claim tokens withdrawal of which was initiated back in late 2022, tokens burned, message stored in 0x42..00 bridge slot, slot is known(I traced it with debug_traceCall method using "OP classic" free node in blockpi, debug_traceTransaction won't work and will always fail with "tracing failed: insufficient funds for gas * price + value: address 0x... have 0 want ...", debug_traceCall with block number -1 of actual is your friend to get the pre/post slot in 0x42....00 contract).

Initially I thought I'll need to use OptimismPortal.proveWithdrawalTransaction, and I've gathered eth_getProof of that slot at that block, gathered latestL2Block block info from the recent "game" contract to populate all fields exactly like it's done nowadays, but this approach didn't work, reverting at OptimismPortal_InvalidOutputRootProof(). I've built it exactly like today's withdrawals do. It's all undocumented basically and very obscured in viem library. I will do a writeup on all this once I'm finished but for now key question is what do I do with this withdrawal? How do I claim that? Using which contract? Back in the day it wasn't OptimismPortal it was L1CrossDomainMessenger.relayMessage(). Am I supposed to be using that contract instead today? If so, how/where do I gather that proof(and proof of what exactly? slot at that block that stored withdrawal message hash? or what?)?

A lot(most) links to older docs and sources are broken some are not in web archive either.

Please help preserve this sacred knowledge and I'll do a full writeup on manual withdrawals, pre-bedrock and post-bedrock, without libraries, all by hand. I think this is going to be invaluable resource for future as while I was looking for sample txes to compare to I found a lot of people who didn't finalize their pre-bedrock withdrawals too. Perhaps due to lack of any docs on how to do it today and bridge apps support.

// I won't be posting my withdrawal txid but can post any other finalized ones for reference.

If you are from OP team I also have an open thread in DC.

Did you check for duplicate questions?

• Yes

Similar Questions or Issues Found

No response

Please select the type of request

• Bug Report
• Feature Request

Steps to Reproduce

No response

Expected vs. Actual Behaviour

No response

Software Versions

No response

Operating System

No response

Hardware Resources

No response

Configuration Files, Startup Flags & Environment Variables

No response

Error Messages / Logs

No response

Feature Description

No response

Purpose and Benefits

No response

Relevant Context or Examples

No response

[opfocus]

https://www.npmjs.com/package/@eth-optimism/sdk
Although the Optimism SDK has been deprecated, I think you can still refer to its implementation.
The relevant code includes:

const args = [ 
    [
        withdrawal.messageNonce,
        withdrawal.sender,
        withdrawal.target,
        withdrawal.value,
        withdrawal.minGasLimit,
        withdrawal.message,
    ],
    proof.l2OutputIndex,
    [
        proof.outputRootProof.version,
        proof.outputRootProof.stateRoot,
        proof.outputRootProof.messagePasserStorageRoot,
        proof.outputRootProof.latestBlockhash,
    ],
    proof.withdrawalProof,
    (opts === null || opts === void 0 ? void 0 : opts.overrides) || {},
];
return this.contracts.l1.OptimismPortal.populateTransaction.proveWithdrawalTransaction(...args);

[lhcadmins]

Okay, as expected when I'm blindly using l2OutputIndex(block number at which withdrawal was called on L2, pre-bedrock) I get out of bound because proveWithdrawalTransaction wants that as _disputeGameIndex

This obviously happens even earlier now, before any of proof validation. I need to figure out what's wrong with proof I'm providing it with.

[Chomtana]

You might have to look at this contract

https://etherscan.io/address/0x25aCe71c97B33Cc4729CF772ae268934F7ab5fA1

To see if you can relay the message

[opfocus]

I spent a lot of time gathering the required parameters and found that many of them can’t be obtained directly — they need complex processing. But if I have to write scripts to handle all this data myself, then what’s the point of not using the SDK in the first place?