feat(store-sync): add support for watching pending logs

[alvrs]

No description provided.

[changeset-bot]

🦋 Changeset detected

Latest commit: 554f484

The changes in this PR will be included in the next version bump.

This PR includes changesets to release 24 packages

Name	Type
@latticexyz/store-sync	Patch
@latticexyz/dev-tools	Patch
@latticexyz/explorer	Patch
@latticexyz/store-indexer	Patch
@latticexyz/abi-ts	Patch
@latticexyz/block-logs-stream	Patch
@latticexyz/cli	Patch
@latticexyz/common	Patch
@latticexyz/config	Patch
create-mud	Patch
@latticexyz/faucet	Patch
@latticexyz/gas-report	Patch
@latticexyz/protocol-parser	Patch
@latticexyz/react	Patch
@latticexyz/recs	Patch
@latticexyz/schema-type	Patch
solhint-config-mud	Patch
solhint-plugin-mud	Patch
@latticexyz/stash	Patch
@latticexyz/store	Patch
@latticexyz/utils	Patch
@latticexyz/world-module-metadata	Patch
@latticexyz/world-modules	Patch
@latticexyz/world	Patch

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

[holic]

I think we need to move around the onProgress handler so that we can communicate to the client when we're hydrated. But I don't think the current watchLogs approach communicates that super well, so unclear how to add this behavior.

[holic]

Suggested change

	if ("error" in response) {
	if ("error" in response) {
	// if error isn't associated with a request ID, ignore
	// TODO: keep track of our request IDs and ignore others?
	if ("id" in response && response.id != null) return;

[holic]

actually could maybe lift this early return outside the error handler?

[holic]

I discovered that something sending other random RPC calls (net_version specifically, so I think it was Metamask) to the websocket will cause this to close

[alvrs]

it should currently open a separate websocket connection for this log stream, did you experience any other errors come through here?

[holic]

yes, I saw net_version occasionally sent through this (and not sure where from or how it got a reference to this open socket, but my only guess is metamask) and that would fail on our node, causing the stream to close, until it was recently added to the whitelist

[holic]

Turns out net_version request comes from viem's ping method used by keepAlive option. The right move was to enable this at our node rather than fixing anything here.

And it would be good if viem keeps net_version as the ping method (rather than e.g. moving to eth_chainId) because we cache eth_chainId requests to reduce end-to-end latency of the transaction request lifecycle.