Specify a placeholder TTD

[paulhauner]

Allowing the successful exchange of the transition config values before the TTD is known allows users to setup their EL+CL today. This helps ensure the network is Ready For The Merge™ and helps with Testing The Merge™.

The consensus specs specify a placeholder TTD of 115792089237316195423570985008687907853269984665640564039457584007913129638912, which is equal to 2**256-2**10. This value is absurdly high, but not the full size of a uint256 such that we can still perform tests with greater-than-TTD values.

AFAIK, there is no equivalent placeholder value in the EL specs. Geth presently returns a TTD of null and Nethermind a value of 0.

For reasons specified earlier, I think CL and EL should agree on these placeholder values. Since a specification already exists for the CL (although it is somewhat implicit), I propose that we explicitly apply the same specification to both EL and CL.

[djrtwo]

I'm okay with this

@paulhauner do you see this as necessary given the mainnet timeline?

[paulhauner]

@paulhauner do you see this as necessary given the mainnet timeline?

I'm not sure about must-have, but I think it's worthwhile. Not having this in the specification has already proved itself to be a mistake; it prevented users from proactively preparing for the merge. Adding this to the spec doesn't seem immediately necessary for mainnet, but I see it as tidying a loose-end and and paying it forward to future testnets and other chains who want to follow the roads we've paved (e.g., gnosis).

Since it's sitting here on a plate in front of us, I'd say we merge it. I certainly won't be heartbroken if it doesn't get merged, though. Definitely not dying on (or feeling even mildly uncomfortable on) this hill.