A place for talking about the Ethereum Virtual Machine on Farcaster.

Logs are the weakest link in eth development
- eth_getLogs is slow, lacks pagination
- eth_createFilter and getFilterLogs are flaky in my experience
- returned logs missing timestamp, userOpHash
- ETH transfer don't emit anything at all
- called "events" interchangeably

Real World Ethereum 🌎 
building /daimo ✳️ 

I give you a list of filters. You give me the following immediately + again as each new block arrives:

{
  fromBlock, toBlock,
  fullLogs: [
    { filterIndex, transactionHash, userOpHash?, indexInOpOrTx, blockHash, blockTimestamp, log?: { existing fields }, ethTransfer?: { from, to, amount } }, ...
  ]
}

Makes it easy to keep running state, notify on events, handle reorgs correctly. Faster indexing.

1 customer and 100 points to whoever makes a new ...FullLogs API fixing the above. 200 if you include websocket push

Future ethereum transactions use erc-4337 account abstraction. These are called “user operations” in AA jargon, and are bundled many userops to one (old-style) transaction.

So if u see a Transfer log and want to know which (modern) tx it came from? Too bad, current API makes this hard.

We solve for this at Goldsky.com.

Blockchain indexing is a pain in the butt, so we built a massively powerful realtime indexer + data pipeline to stream everything emitted onchain anywhere you need it.

Super easy to create subgraphs + flexible enough to pull decoded log / transaction data that is actually meaningful.