feat(rpc): migrate base_subscribe to eth_subscribe #272

madisoncarter1234 · 2025-12-10T14:43:07Z

Summary

Migrates flashblocks subscriptions from the base namespace to the eth namespace per #250.

Since no release has been cut with base_subscribe yet, this is a clean replacement with no backwards compatibility needed.

Rename BasePubSubApi → EthPubSubApi (namespace: base → eth)
Rename BasePubSub → EthPubSub
Rename BaseSubscriptionKind → EthSubscriptionKind
Use replace_configured to override reth's existing eth_subscribe
Update all tests to use eth_subscribe/eth_unsubscribe

Usage

- method: "base_subscribe", params: ["newFlashblocks"]
+ method: "eth_subscribe", params: ["newFlashblocks"]

- method: "base_unsubscribe"
+ method: "eth_unsubscribe"

Test plan

All 18 flashblocks RPC tests pass
Release build succeeds

Closes #250

cb-heimdall · 2025-12-10T14:43:12Z

✅ Heimdall Review Status

Requirement Status More Info

Reviews

✅ 1/1

Denominator calculation

Show calculation

1 if user is bot 0

1 if user is external 0

2 if repo is sensitive 0

From .codeflow.yml 1

Additional review requirements

Show calculation

Max	0

0

From CODEOWNERS 0

Global minimum 0

Max 1

1

1 if commit is unverified 1

Sum 2

Migrates flashblocks subscriptions from the base namespace to eth namespace per issue base#250. Since no release has been cut with base_subscribe yet, this is a clean replacement with no backwards compatibility needed. Changes: - Rename BasePubSubApi -> EthPubSubApi (namespace: base -> eth) - Rename BasePubSub -> EthPubSub - Rename BaseSubscriptionKind -> EthSubscriptionKind - Use replace_configured to override reth's existing eth_subscribe - Update all tests to use eth_subscribe/eth_unsubscribe Closes base#250

refcell

This looks good to me. Would appreciate second set of eyes from @danyalprout. Since we're using replace_configured on the module context, we shouldn't run into any namespace collision errors and the test suite should validate this behavior. Nice work @madisoncarter1234!

madisoncarter1234 · 2025-12-11T19:48:02Z

This looks good to me. Would appreciate second set of eyes from @danyalprout. Since we're using replace_configured on the module context, we shouldn't run into any namespace collision errors and the test suite should validate this behavior. Nice work @madisoncarter1234!

glad to help!!

danyalprout · 2025-12-11T20:35:28Z

This looks great re: renaming. As we're overriding, do we need to extend the enum for the standard subscription types and proxy them to Reth?

madisoncarter1234 · 2025-12-11T21:26:09Z

This looks great re: renaming. As we're overriding, do we need to extend the enum for the standard subscription types and proxy them to Reth?

Right.‍‌‍‍‌‍‌‍‍‌ The initial PR was mainly about the namespace migration from base_subscribe to eth_subscribe and didn't really consider the full enum extension that you mentioned in #250. Because replace_configured changes the whole eth_subscribe handler, standard subscription types such as newHeads and logs would stop working if used with the current implementation.

I will change the enum to include standard types and proxy them to reth's underlying handler. A revision will be ‍‌‍‍‌‍‌‍‍‌pushed.

danyalprout · 2025-12-11T22:01:29Z

Amazing -- thank you @madisoncarter1234

Per base#250, this extends the eth_subscribe handler to support both standard Ethereum subscription types and Base-specific flashblocks subscriptions. Changes: - Rename EthSubscriptionKind to ExtendedSubscriptionKind - Add standard types: newHeads, logs, newPendingTransactions, syncing - Add Base-specific types: newFlashblocks, pendingLogs - Proxy standard subscription types to reth's underlying implementation - Add reth-rpc dependency for access to RethEthPubSub - Update EthPubSub to accept eth_api for proxying - Add test for standard newHeads subscription This ensures backwards compatibility with existing eth_subscribe usage while adding flashblocks-specific subscription capabilities.

madisoncarter1234 · 2025-12-12T01:45:38Z

Updated‍‌‍‍‌‍‌‍‍‌ the implementation to extend eth_subscribe with full support for standard subscription types per #250.

Changes:

Renamed EthSubscriptionKind to ExtendedSubscriptionKind
Added standard types: newHeads, logs, newPendingTransactions, syncing
Added Base-specific types: newFlashblocks, pendingLogs
Standard subscription types are now proxied to reth's underlying EthPubSub implementation
Added reth-rpc dependency for access to reth's pubsub handler

This maintains backwards compatibility, existing eth_subscribe("newHeads") calls will still function as they have been while also allowing for flashblocks-specific subscriptions to be used.

Regarding the overriding of newPendingTransactions to use flashblocks state, I have left it as standard behavior for now since it has different semantics (mempool txs vs sequenced txs) I would be glad to create a flashblocks specific variant in a follow up if you want

refcell

This looks great; nice work!

haardikk21 · 2025-12-12T15:04:26Z

crates/rpc/src/base/types.rs

-    /// New flashblocks subscription.
+pub enum ExtendedSubscriptionKind {
+    /// New block headers subscription (standard).
+    NewHeads,


nit: maybe this enum should encapsulate the SubscriptionKind enum rather than redefining it. while i dont expect the signature to change anytime soon, if it ever were to change, it would be nice to automatically have support for it or automatically get an unhandled error

➕ I like this approach as well.

haardikk21 · 2025-12-12T15:05:54Z

crates/rpc/tests/flashblocks_rpc.rs

+    assert_eq!(sub["jsonrpc"], "2.0");
+    assert_eq!(sub["id"], 1);
+    // Should return a subscription ID, confirming the subscription was accepted
+    assert!(sub["result"].is_string(), "Expected subscription ID, got: {:?}", sub);


is is_string the correct assertion to make here?

danyalprout

This looks great, I'm happy to experiment with the enum proposal in a follow up.

madisoncarter1234 · 2025-12-12T21:52:13Z

Thanks everyone!!

madisoncarter1234 mentioned this pull request Dec 10, 2025

Migrate base_subscribe to eth_subscribe #250

Closed

madisoncarter1234 force-pushed the feat/migrate-base-subscribe-to-eth branch from 0efce22 to 58c4468 Compare December 10, 2025 22:56

refcell approved these changes Dec 10, 2025

View reviewed changes

madisoncarter1234 requested a review from refcell December 12, 2025 01:40

refcell approved these changes Dec 12, 2025

View reviewed changes

chore(ci): fix lints

f1cc724

haardikk21 reviewed Dec 12, 2025

View reviewed changes

danyalprout approved these changes Dec 12, 2025

View reviewed changes

danyalprout merged commit 08bd601 into base:main Dec 12, 2025
10 checks passed

refcell mentioned this pull request Dec 12, 2025

chore(rpc): Properly Extend Subscription Kind #277

Merged

madisoncarter1234 mentioned this pull request Dec 15, 2025

feat(rpc): Add flashblocks-aware newPendingTransactions subscription variant #280

Closed

feat(rpc): migrate base_subscribe to eth_subscribe #272

feat(rpc): migrate base_subscribe to eth_subscribe #272

Uh oh!

Conversation

madisoncarter1234 commented Dec 10, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Usage

Test plan

Uh oh!

cb-heimdall commented Dec 10, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

✅ Heimdall Review Status

Uh oh!

refcell left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

madisoncarter1234 commented Dec 11, 2025

Uh oh!

danyalprout commented Dec 11, 2025

Uh oh!

madisoncarter1234 commented Dec 11, 2025

Uh oh!

danyalprout commented Dec 11, 2025

Uh oh!

madisoncarter1234 commented Dec 12, 2025

Uh oh!

refcell left a comment

Choose a reason for hiding this comment

Uh oh!

haardikk21 Dec 12, 2025

Choose a reason for hiding this comment

Uh oh!

refcell Dec 12, 2025

Choose a reason for hiding this comment

Uh oh!

haardikk21 Dec 12, 2025

Choose a reason for hiding this comment

Uh oh!

danyalprout left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

madisoncarter1234 commented Dec 12, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

madisoncarter1234 commented Dec 10, 2025 •

edited

Loading

cb-heimdall commented Dec 10, 2025 •

edited

Loading

refcell left a comment •

edited

Loading