Proposal NEARWEEK Upgrades for near-staking.com

[trechriron]

https://docs.google.com/document/d/1AGoDUH1WWjN_-duOmUbr69vjmBkarpWmxE13IQ2TvvU/edit?usp=sharing

This thread is for the discussion of the following proposal.

[trechriron]

Your proposal has been tentatively approved. The Infrastructure Committee requests the following modifications.

1. The committee does not recommend that you establish your own RPC. Instead, we recommend researching and budgeting for services from existing providers in the NEAR space. The initial grant can assist in covering these costs.
2. We would like you to thoroughly research and include the cost and effort of integrating a wallet-neutral staking interface in your proposal. This is a crucial aspect that needs to be addressed.
   1. Please contact (TG handle forthcoming…) at Pagoda for assistance, who has already started work on an open-source staking interface.
   2. You should also be mindful of the issues with the rewards calculation. It is not always accurate and can abuse calls to RPC nodes. We would like your proposal for a path forward to address these issues.

We encourage you to ask any questions or seek clarification on any points.

[AngelBlock]

Hi @trechriron, thanks for the reply and apologies for late reply, we have been discussing internally.

Please see my answers below:

The committee does not recommend that you establish your own RPC. Instead, we recommend researching and budgeting for services from existing providers in the NEAR space. The initial grant can assist in covering these costs.

For optimal performance and stability of near-staking, it is crucial to use a local RPC node to ensure low latency and high performance. Utilizing an external RPC for most of the data is not feasible due to the high volume of data pulled frequently, including information on all delegators and validators.

However, for less intensive calls, we use a mainnet-archival RPC. We are interested in finding an existing provider that can offer similar or better performance/rates than current mannet archival RPC nodes. Can you please recommend any providers?

We would like you to thoroughly research and include the cost and effort of integrating a wallet-neutral staking interface in your proposal. This is a crucial aspect that needs to be addressed.
Please contact (TG handle forthcoming…) at Pagoda for assistance, who has already started work on an open-source staking interface.

This topic has already been discussed with the suggested resource. However, since they are still in the early stages, we can only provide a rough estimate of the integration costs without knowing all the details.

You should also be mindful of the issues with the rewards calculation. It is not always accurate and can abuse calls to RPC nodes. We would like your proposal for a path forward to address these issues.

We are aware of minor calculation issues in some cases due to gaps in data. Since the archival RPCs also lack this data, we have been unable to identify a resource to fill these gaps. Additionally, there are some imperfections in the calculation algorithm for certain edge cases, affecting no more than 0.5% of all NEAR accounts. This is partly why we are seeking funding to improve the algorithm. It requires thorough QA, as most of the edge cases are not obvious and extensive testing is needed.

Could you please clarify what you mean by "abuse calls to RPC nodes"? We do not use external RPC for any data involved in rewards calculation, so there is no risk of abuse in that context. The only feature where we use external RPC is for retrieving account balances, which is not resource-intensive and, according to our analytics, is barely used.

I hope this clarifies your questions, and we can proceed further.