`CheckPoint` design issue

[ValuedMammal]

These are some of the proposed changes to CheckPoint. I'm opening this as a record of ongoing discussion as it relates to the CheckPoint structure and functionality.

• Make CheckPoint data field optional feat(chain)!: make CheckPoint data field optional #2024
• Validate CheckPoint links via previous hash Ensure CheckPoint chain methods validate and link via previous blockhash #2021
• Add MTP calculation Add median-time-past (MTP) calculation to CheckPoint #2036
• Add efficient comparison Add Efficient Sparse Chain Comparison to CheckPoint #2033
• Improve search complexity Add skiplist to CheckPoint for O(√n) traversal #2048

[ValuedMammal]

In order to evaluate something proposing a significant change to the structure or design of the CheckPoint, it's important to clarify the specific requirements of the CheckPoint. What does this data structure need to accomplish, and what is the best way to do it while keeping the tradeoffs in mind?

• If there are enhancements or optimizations to be made, can they be done within the current design contraints?
• If it's a design change being proposed, what is the new design? Does it uphold the same requirements?
• Is there a chance to simplify by benefiting from types that exist either in the std library or elsewhere?
• Is there a sufficiently motivated change that warrants the time and investment on a complete rewrite?

[oleonardolima]

Awesome, thanks for compiling all the issues here! Agree, I think it's a good idea to have the big picture and the tradeoffs clear in mind. I'd add that we should probably compile all these changes in a "Checkpoint's ADR" at some point.

I'll try to come up with the answers from my perspective once I settle back home later.