project proposal - #192
Conversation
taxmeifyoucan
left a comment
taxmeifyoucan
left a comment
There was a problem hiding this comment.
I did not finish the review because of the main issue with the document - it's just too long and heavy. Proposals need to be digestible, understandable by human, mentors. This feels very AI generated, too verbose, you should spend time on writing deliverables with your own words instead. It's also very ambitious, complex project that might require multiple people to work on this
| * state that may exist but was intentionally not retained; | ||
| * state tequired for execution but currently unavailable. | ||
|
|
||
| This distinctions affects Reth's storage providers, snap synchronization, RPC methods, EVM execution, transaction-pool validation, Engine API behaviour, reorganization handling, and future support for state expiry. |
There was a problem hiding this comment.
This is getting into project description/specs, keep the motivation brief. In the paragraph above, you can just shortly introduce statelessness and what makes the reth implementation unique
There was a problem hiding this comment.
also the motivation more pragmatically, db size reduced eg in geth ~640 GiB to ~59 GiB
| * unavailable state RPC semantics; | ||
| * the assumption Reth currently makes about complete local state. | ||
|
|
||
| For every account, the node retains account-level information such as: |
There was a problem hiding this comment.
too many bullet points that could be better fit in specs.. in this section focus more on which parts of the client are affected and high level overview of the implementation
| | Stretch work | Txpool policy, Engine API behavior, dynamic retention filters, deeper reorganization recovery | | ||
|
|
||
| ## Research Questions | ||
|
|
There was a problem hiding this comment.
I don't really understand need for 7 abstract research questions with no connection to specific code paths. If there are open questions, they should be connected to specific deliverables, be grounded in the code..
2 participants
No description provided.