Increase Effect RPC ping timeout for large snapshots

[Pearcekieser]

What Changed

Increases the patched Effect RPC pinger delay from 5 seconds to 60 seconds.

This changes the existing patches/effect@4.0.0-beta.73.patch so the generated RPC client waits longer before declaring the connection unhealthy due to a missed Pong.

Why

Large thread snapshots can occupy the WebSocket long enough that the Effect RPC Pong is queued behind snapshot chunks. On slow remote connections, the current 5 second heartbeat window can expire before the Pong reaches the client, causing a reconnect. The reconnect then requests the same large snapshot again, which can repeat the failure loop.

A 60 second window gives large snapshot sends more time to drain before the client treats the connection as dead.

This is a small mitigation for #2761. It does not solve the deeper issue that large snapshots and control/heartbeat messages share the same WebSocket queue; longer term, snapshots likely need pagination, truncation, backpressure, or control-message prioritization.

Checklist

• This PR is small and focused
• I explained what changed and why
• UI changes are not applicable

Note

Increase Effect RPC ping interval to 60 seconds for large snapshot support

• Updates patches/effect@4.0.0-beta.78.patch to change the protocol worker ping interval from 5 seconds to 60 seconds.
• Adds an hooks.onPing call (when provided) before each ping is sent, allowing pre-ping side effects.
• Behavioral Change: ping frequency drops from every 5s to every 60s, which may affect connection keepalive behavior in environments with short idle timeouts.

^{Macroscope summarized 2825718.}

---

Note

Medium Risk
Changes connection health detection for all Effect RPC WebSocket clients; a 60s window delays detection of truly dead peers but reduces false reconnect loops during heavy snapshot traffic.

Overview
Updates the vendored effect@4.0.0-beta.78 patch so the RPC socket pinger waits 60 seconds between ping attempts instead of 5 seconds, giving large WebSocket transfers (e.g. thread snapshots) more time before the client treats a missing Pong as a dead connection.

The same patch delta wires ConnectionHooks into the ping loop: onPong runs after recording a pong, onPing runs before writePing, and both default to Effect.void when unset. pnpm-lock.yaml is refreshed so every workspace consumer resolves the new effect patch hash.

^{Reviewed by Cursor Bugbot for commit 2825718. Bugbot is set up for automated code reviews on this repo. Configure here.}

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: 869b0f55-01e6-4fd4-8ceb-cad2894b8ac5

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[macroscopeapp]

Approvability

Verdict: Approved

Simple timeout configuration change from 5 to 60 seconds in an Effect RPC patch to handle large snapshots. The lockfile changes are mechanical hash updates from the patch modification.

^{You can customize Macroscope's approvability policy. Learn more.}

[t3dotgg]

I resolved the rebase locally, but this PR’s branch is in the organization-owned pearce-codes/t3code fork. GitHub rejects maintainer pushes to that fork, and the Update Branch API also reports a conflict. Please rebase the branch onto current main (the patch now targets patches/effect@4.0.0-beta.78.patch) or grant push access to the fork branch.

Dismissing prior approval to re-evaluate 6e69066

Dismissing prior approval to re-evaluate 644cc92

[Pearcekieser]

Thank for looking at this. Sorry for the delay was out on vacation. I recently started running into this timeouts again, so I'm hopeful this will help me and others.

I updated this to main. I don't think orgs owned branches allow pushes from upstream's maintainers, so I re opened this under a different PR if you need to edit.
New PR on a maintainer-pushable branch: #3817

Also @mincua suggested the same fix here so please try to give him credit if you merge this: #3500

Seems like this could also help fix:
#3553
#3734

Dismissing prior approval to re-evaluate 2825718

[juliusmarminge]

Closing because the author-designated maintainer-editable replacement is #3817. The replacement PR explicitly carries this work forward.

[Pearcekieser]

What Changed

Increases the patched Effect RPC pinger delay from 5 seconds to 60 seconds.

This changes the existing patches/effect@4.0.0-beta.73.patch so the generated RPC client waits longer before declaring the connection unhealthy due to a missed Pong.

Why

Large thread snapshots can occupy the WebSocket long enough that the Effect RPC Pong is queued behind snapshot chunks. On slow remote connections, the current 5 second heartbeat window can expire before the Pong reaches the client, causing a reconnect. The reconnect then requests the same large snapshot again, which can repeat the failure loop.

A 60 second window gives large snapshot sends more time to drain before the client treats the connection as dead.

This is a small mitigation for #2761. It does not solve the deeper issue that large snapshots and control/heartbeat messages share the same WebSocket queue; longer term, snapshots likely need pagination, truncation, backpressure, or control-message prioritization.

Checklist

• This PR is small and focused
• I explained what changed and why
• UI changes are not applicable

Note

Increase Effect RPC ping interval to 60 seconds for large snapshot support

• Updates patches/effect@4.0.0-beta.78.patch to change the protocol worker ping interval from 5 seconds to 60 seconds.
• Adds an hooks.onPing call (when provided) before each ping is sent, allowing pre-ping side effects.
• Behavioral Change: ping frequency drops from every 5s to every 60s, which may affect connection keepalive behavior in environments with short idle timeouts.

^{Macroscope summarized 2825718.}

---

Note

Medium Risk
Changes connection health detection for all Effect RPC WebSocket clients; a 60s window delays detection of truly dead peers but reduces false reconnect loops during heavy snapshot traffic.

Overview
Updates the vendored effect@4.0.0-beta.78 patch so the RPC socket pinger waits 60 seconds between ping attempts instead of 5 seconds, giving large WebSocket transfers (e.g. thread snapshots) more time before the client treats a missing Pong as a dead connection.

The same patch delta wires ConnectionHooks into the ping loop: onPong runs after recording a pong, onPing runs before writePing, and both default to Effect.void when unset. pnpm-lock.yaml is refreshed so every workspace consumer resolves the new effect patch hash.

^{Reviewed by Cursor Bugbot for commit 2825718. Bugbot is set up for automated code reviews on this repo. Configure here.}

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: 869b0f55-01e6-4fd4-8ceb-cad2894b8ac5

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[macroscopeapp]

Approvability

Verdict: Approved

Simple timeout configuration change from 5 to 60 seconds in an Effect RPC patch to handle large snapshots. The lockfile changes are mechanical hash updates from the patch modification.

^{You can customize Macroscope's approvability policy. Learn more.}

[t3dotgg]

I resolved the rebase locally, but this PR’s branch is in the organization-owned pearce-codes/t3code fork. GitHub rejects maintainer pushes to that fork, and the Update Branch API also reports a conflict. Please rebase the branch onto current main (the patch now targets patches/effect@4.0.0-beta.78.patch) or grant push access to the fork branch.

Dismissing prior approval to re-evaluate 6e69066

Dismissing prior approval to re-evaluate 644cc92

[Pearcekieser]

Thank for looking at this. Sorry for the delay was out on vacation. I recently started running into this timeouts again, so I'm hopeful this will help me and others.

I updated this to main. I don't think orgs owned branches allow pushes from upstream's maintainers, so I re opened this under a different PR if you need to edit.
New PR on a maintainer-pushable branch: #3817

Also @mincua suggested the same fix here so please try to give him credit if you merge this: #3500

Seems like this could also help fix:
#3553
#3734

Dismissing prior approval to re-evaluate 2825718

[juliusmarminge]

Closing because the author-designated maintainer-editable replacement is #3817. The replacement PR explicitly carries this work forward.

[Pearcekieser]

What Changed

Increases the patched Effect RPC pinger delay from 5 seconds to 60 seconds.

This changes the existing patches/effect@4.0.0-beta.73.patch so the generated RPC client waits longer before declaring the connection unhealthy due to a missed Pong.

Why

Large thread snapshots can occupy the WebSocket long enough that the Effect RPC Pong is queued behind snapshot chunks. On slow remote connections, the current 5 second heartbeat window can expire before the Pong reaches the client, causing a reconnect. The reconnect then requests the same large snapshot again, which can repeat the failure loop.

A 60 second window gives large snapshot sends more time to drain before the client treats the connection as dead.

This is a small mitigation for #2761. It does not solve the deeper issue that large snapshots and control/heartbeat messages share the same WebSocket queue; longer term, snapshots likely need pagination, truncation, backpressure, or control-message prioritization.

Checklist

• This PR is small and focused
• I explained what changed and why
• UI changes are not applicable

Note

Increase Effect RPC ping interval to 60 seconds for large snapshot support

• Updates patches/effect@4.0.0-beta.78.patch to change the protocol worker ping interval from 5 seconds to 60 seconds.
• Adds an hooks.onPing call (when provided) before each ping is sent, allowing pre-ping side effects.
• Behavioral Change: ping frequency drops from every 5s to every 60s, which may affect connection keepalive behavior in environments with short idle timeouts.

^{Macroscope summarized 2825718.}

---

Note

Medium Risk
Changes connection health detection for all Effect RPC WebSocket clients; a 60s window delays detection of truly dead peers but reduces false reconnect loops during heavy snapshot traffic.

Overview
Updates the vendored effect@4.0.0-beta.78 patch so the RPC socket pinger waits 60 seconds between ping attempts instead of 5 seconds, giving large WebSocket transfers (e.g. thread snapshots) more time before the client treats a missing Pong as a dead connection.

The same patch delta wires ConnectionHooks into the ping loop: onPong runs after recording a pong, onPing runs before writePing, and both default to Effect.void when unset. pnpm-lock.yaml is refreshed so every workspace consumer resolves the new effect patch hash.

^{Reviewed by Cursor Bugbot for commit 2825718. Bugbot is set up for automated code reviews on this repo. Configure here.}

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: 869b0f55-01e6-4fd4-8ceb-cad2894b8ac5

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[macroscopeapp]

Approvability

Verdict: Approved

Simple timeout configuration change from 5 to 60 seconds in an Effect RPC patch to handle large snapshots. The lockfile changes are mechanical hash updates from the patch modification.

^{You can customize Macroscope's approvability policy. Learn more.}

[t3dotgg]

I resolved the rebase locally, but this PR’s branch is in the organization-owned pearce-codes/t3code fork. GitHub rejects maintainer pushes to that fork, and the Update Branch API also reports a conflict. Please rebase the branch onto current main (the patch now targets patches/effect@4.0.0-beta.78.patch) or grant push access to the fork branch.

Dismissing prior approval to re-evaluate 6e69066

Dismissing prior approval to re-evaluate 644cc92

[Pearcekieser]

Thank for looking at this. Sorry for the delay was out on vacation. I recently started running into this timeouts again, so I'm hopeful this will help me and others.

I updated this to main. I don't think orgs owned branches allow pushes from upstream's maintainers, so I re opened this under a different PR if you need to edit.
New PR on a maintainer-pushable branch: #3817

Also @mincua suggested the same fix here so please try to give him credit if you merge this: #3500

Seems like this could also help fix:
#3553
#3734

Dismissing prior approval to re-evaluate 2825718

[juliusmarminge]

Closing because the author-designated maintainer-editable replacement is #3817. The replacement PR explicitly carries this work forward.

[Pearcekieser]

What Changed

Increases the patched Effect RPC pinger delay from 5 seconds to 60 seconds.

This changes the existing patches/effect@4.0.0-beta.73.patch so the generated RPC client waits longer before declaring the connection unhealthy due to a missed Pong.

Why

Large thread snapshots can occupy the WebSocket long enough that the Effect RPC Pong is queued behind snapshot chunks. On slow remote connections, the current 5 second heartbeat window can expire before the Pong reaches the client, causing a reconnect. The reconnect then requests the same large snapshot again, which can repeat the failure loop.

A 60 second window gives large snapshot sends more time to drain before the client treats the connection as dead.

This is a small mitigation for #2761. It does not solve the deeper issue that large snapshots and control/heartbeat messages share the same WebSocket queue; longer term, snapshots likely need pagination, truncation, backpressure, or control-message prioritization.

Checklist

• This PR is small and focused
• I explained what changed and why
• UI changes are not applicable

Note

Increase Effect RPC ping interval to 60 seconds for large snapshot support

• Updates patches/effect@4.0.0-beta.78.patch to change the protocol worker ping interval from 5 seconds to 60 seconds.
• Adds an hooks.onPing call (when provided) before each ping is sent, allowing pre-ping side effects.
• Behavioral Change: ping frequency drops from every 5s to every 60s, which may affect connection keepalive behavior in environments with short idle timeouts.

^{Macroscope summarized 2825718.}

---

Note

Medium Risk
Changes connection health detection for all Effect RPC WebSocket clients; a 60s window delays detection of truly dead peers but reduces false reconnect loops during heavy snapshot traffic.

Overview
Updates the vendored effect@4.0.0-beta.78 patch so the RPC socket pinger waits 60 seconds between ping attempts instead of 5 seconds, giving large WebSocket transfers (e.g. thread snapshots) more time before the client treats a missing Pong as a dead connection.

The same patch delta wires ConnectionHooks into the ping loop: onPong runs after recording a pong, onPing runs before writePing, and both default to Effect.void when unset. pnpm-lock.yaml is refreshed so every workspace consumer resolves the new effect patch hash.

^{Reviewed by Cursor Bugbot for commit 2825718. Bugbot is set up for automated code reviews on this repo. Configure here.}

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: 869b0f55-01e6-4fd4-8ceb-cad2894b8ac5

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[macroscopeapp]

Approvability

Verdict: Approved

Simple timeout configuration change from 5 to 60 seconds in an Effect RPC patch to handle large snapshots. The lockfile changes are mechanical hash updates from the patch modification.

^{You can customize Macroscope's approvability policy. Learn more.}

[t3dotgg]

I resolved the rebase locally, but this PR’s branch is in the organization-owned pearce-codes/t3code fork. GitHub rejects maintainer pushes to that fork, and the Update Branch API also reports a conflict. Please rebase the branch onto current main (the patch now targets patches/effect@4.0.0-beta.78.patch) or grant push access to the fork branch.

Dismissing prior approval to re-evaluate 6e69066

Dismissing prior approval to re-evaluate 644cc92

[Pearcekieser]

Thank for looking at this. Sorry for the delay was out on vacation. I recently started running into this timeouts again, so I'm hopeful this will help me and others.

I updated this to main. I don't think orgs owned branches allow pushes from upstream's maintainers, so I re opened this under a different PR if you need to edit.
New PR on a maintainer-pushable branch: #3817

Also @mincua suggested the same fix here so please try to give him credit if you merge this: #3500

Seems like this could also help fix:
#3553
#3734

Dismissing prior approval to re-evaluate 2825718

[juliusmarminge]

Closing because the author-designated maintainer-editable replacement is #3817. The replacement PR explicitly carries this work forward.

[Pearcekieser]

What Changed

Increases the patched Effect RPC pinger delay from 5 seconds to 60 seconds.

This changes the existing patches/effect@4.0.0-beta.73.patch so the generated RPC client waits longer before declaring the connection unhealthy due to a missed Pong.

Why

Large thread snapshots can occupy the WebSocket long enough that the Effect RPC Pong is queued behind snapshot chunks. On slow remote connections, the current 5 second heartbeat window can expire before the Pong reaches the client, causing a reconnect. The reconnect then requests the same large snapshot again, which can repeat the failure loop.

A 60 second window gives large snapshot sends more time to drain before the client treats the connection as dead.

This is a small mitigation for #2761. It does not solve the deeper issue that large snapshots and control/heartbeat messages share the same WebSocket queue; longer term, snapshots likely need pagination, truncation, backpressure, or control-message prioritization.

Checklist

• This PR is small and focused
• I explained what changed and why
• UI changes are not applicable

Note

Increase Effect RPC ping interval to 60 seconds for large snapshot support

• Updates patches/effect@4.0.0-beta.78.patch to change the protocol worker ping interval from 5 seconds to 60 seconds.
• Adds an hooks.onPing call (when provided) before each ping is sent, allowing pre-ping side effects.
• Behavioral Change: ping frequency drops from every 5s to every 60s, which may affect connection keepalive behavior in environments with short idle timeouts.

^{Macroscope summarized 2825718.}

---

Note

Medium Risk
Changes connection health detection for all Effect RPC WebSocket clients; a 60s window delays detection of truly dead peers but reduces false reconnect loops during heavy snapshot traffic.

Overview
Updates the vendored effect@4.0.0-beta.78 patch so the RPC socket pinger waits 60 seconds between ping attempts instead of 5 seconds, giving large WebSocket transfers (e.g. thread snapshots) more time before the client treats a missing Pong as a dead connection.

The same patch delta wires ConnectionHooks into the ping loop: onPong runs after recording a pong, onPing runs before writePing, and both default to Effect.void when unset. pnpm-lock.yaml is refreshed so every workspace consumer resolves the new effect patch hash.

^{Reviewed by Cursor Bugbot for commit 2825718. Bugbot is set up for automated code reviews on this repo. Configure here.}

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: 869b0f55-01e6-4fd4-8ceb-cad2894b8ac5

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[macroscopeapp]

Approvability

Verdict: Approved

Simple timeout configuration change from 5 to 60 seconds in an Effect RPC patch to handle large snapshots. The lockfile changes are mechanical hash updates from the patch modification.

^{You can customize Macroscope's approvability policy. Learn more.}

[t3dotgg]

I resolved the rebase locally, but this PR’s branch is in the organization-owned pearce-codes/t3code fork. GitHub rejects maintainer pushes to that fork, and the Update Branch API also reports a conflict. Please rebase the branch onto current main (the patch now targets patches/effect@4.0.0-beta.78.patch) or grant push access to the fork branch.

Dismissing prior approval to re-evaluate 6e69066

Dismissing prior approval to re-evaluate 644cc92

[Pearcekieser]

Thank for looking at this. Sorry for the delay was out on vacation. I recently started running into this timeouts again, so I'm hopeful this will help me and others.

I updated this to main. I don't think orgs owned branches allow pushes from upstream's maintainers, so I re opened this under a different PR if you need to edit.
New PR on a maintainer-pushable branch: #3817

Also @mincua suggested the same fix here so please try to give him credit if you merge this: #3500

Seems like this could also help fix:
#3553
#3734

Dismissing prior approval to re-evaluate 2825718

[juliusmarminge]

Closing because the author-designated maintainer-editable replacement is #3817. The replacement PR explicitly carries this work forward.

[Pearcekieser]

What Changed

Increases the patched Effect RPC pinger delay from 5 seconds to 60 seconds.

This changes the existing patches/effect@4.0.0-beta.73.patch so the generated RPC client waits longer before declaring the connection unhealthy due to a missed Pong.

Why

Large thread snapshots can occupy the WebSocket long enough that the Effect RPC Pong is queued behind snapshot chunks. On slow remote connections, the current 5 second heartbeat window can expire before the Pong reaches the client, causing a reconnect. The reconnect then requests the same large snapshot again, which can repeat the failure loop.

A 60 second window gives large snapshot sends more time to drain before the client treats the connection as dead.

This is a small mitigation for #2761. It does not solve the deeper issue that large snapshots and control/heartbeat messages share the same WebSocket queue; longer term, snapshots likely need pagination, truncation, backpressure, or control-message prioritization.

Checklist

• This PR is small and focused
• I explained what changed and why
• UI changes are not applicable

Note

Increase Effect RPC ping interval to 60 seconds for large snapshot support

• Updates patches/effect@4.0.0-beta.78.patch to change the protocol worker ping interval from 5 seconds to 60 seconds.
• Adds an hooks.onPing call (when provided) before each ping is sent, allowing pre-ping side effects.
• Behavioral Change: ping frequency drops from every 5s to every 60s, which may affect connection keepalive behavior in environments with short idle timeouts.

^{Macroscope summarized 2825718.}

---

Note

Medium Risk
Changes connection health detection for all Effect RPC WebSocket clients; a 60s window delays detection of truly dead peers but reduces false reconnect loops during heavy snapshot traffic.

Overview
Updates the vendored effect@4.0.0-beta.78 patch so the RPC socket pinger waits 60 seconds between ping attempts instead of 5 seconds, giving large WebSocket transfers (e.g. thread snapshots) more time before the client treats a missing Pong as a dead connection.

The same patch delta wires ConnectionHooks into the ping loop: onPong runs after recording a pong, onPing runs before writePing, and both default to Effect.void when unset. pnpm-lock.yaml is refreshed so every workspace consumer resolves the new effect patch hash.

^{Reviewed by Cursor Bugbot for commit 2825718. Bugbot is set up for automated code reviews on this repo. Configure here.}

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: 869b0f55-01e6-4fd4-8ceb-cad2894b8ac5

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[macroscopeapp]

Approvability

Verdict: Approved

Simple timeout configuration change from 5 to 60 seconds in an Effect RPC patch to handle large snapshots. The lockfile changes are mechanical hash updates from the patch modification.

^{You can customize Macroscope's approvability policy. Learn more.}

[t3dotgg]

I resolved the rebase locally, but this PR’s branch is in the organization-owned pearce-codes/t3code fork. GitHub rejects maintainer pushes to that fork, and the Update Branch API also reports a conflict. Please rebase the branch onto current main (the patch now targets patches/effect@4.0.0-beta.78.patch) or grant push access to the fork branch.

Dismissing prior approval to re-evaluate 6e69066

Dismissing prior approval to re-evaluate 644cc92

[Pearcekieser]

Thank for looking at this. Sorry for the delay was out on vacation. I recently started running into this timeouts again, so I'm hopeful this will help me and others.

I updated this to main. I don't think orgs owned branches allow pushes from upstream's maintainers, so I re opened this under a different PR if you need to edit.
New PR on a maintainer-pushable branch: #3817

Also @mincua suggested the same fix here so please try to give him credit if you merge this: #3500

Seems like this could also help fix:
#3553
#3734

Dismissing prior approval to re-evaluate 2825718

[juliusmarminge]

Closing because the author-designated maintainer-editable replacement is #3817. The replacement PR explicitly carries this work forward.

[Pearcekieser]

What Changed

Increases the patched Effect RPC pinger delay from 5 seconds to 60 seconds.

This changes the existing patches/effect@4.0.0-beta.73.patch so the generated RPC client waits longer before declaring the connection unhealthy due to a missed Pong.

Why

Large thread snapshots can occupy the WebSocket long enough that the Effect RPC Pong is queued behind snapshot chunks. On slow remote connections, the current 5 second heartbeat window can expire before the Pong reaches the client, causing a reconnect. The reconnect then requests the same large snapshot again, which can repeat the failure loop.

A 60 second window gives large snapshot sends more time to drain before the client treats the connection as dead.

This is a small mitigation for #2761. It does not solve the deeper issue that large snapshots and control/heartbeat messages share the same WebSocket queue; longer term, snapshots likely need pagination, truncation, backpressure, or control-message prioritization.

Checklist

• This PR is small and focused
• I explained what changed and why
• UI changes are not applicable

Note

Increase Effect RPC ping interval to 60 seconds for large snapshot support

• Updates patches/effect@4.0.0-beta.78.patch to change the protocol worker ping interval from 5 seconds to 60 seconds.
• Adds an hooks.onPing call (when provided) before each ping is sent, allowing pre-ping side effects.
• Behavioral Change: ping frequency drops from every 5s to every 60s, which may affect connection keepalive behavior in environments with short idle timeouts.

^{Macroscope summarized 2825718.}

---

Note

Medium Risk
Changes connection health detection for all Effect RPC WebSocket clients; a 60s window delays detection of truly dead peers but reduces false reconnect loops during heavy snapshot traffic.

Overview
Updates the vendored effect@4.0.0-beta.78 patch so the RPC socket pinger waits 60 seconds between ping attempts instead of 5 seconds, giving large WebSocket transfers (e.g. thread snapshots) more time before the client treats a missing Pong as a dead connection.

The same patch delta wires ConnectionHooks into the ping loop: onPong runs after recording a pong, onPing runs before writePing, and both default to Effect.void when unset. pnpm-lock.yaml is refreshed so every workspace consumer resolves the new effect patch hash.

^{Reviewed by Cursor Bugbot for commit 2825718. Bugbot is set up for automated code reviews on this repo. Configure here.}

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: 869b0f55-01e6-4fd4-8ceb-cad2894b8ac5

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[macroscopeapp]

Approvability

Verdict: Approved

Simple timeout configuration change from 5 to 60 seconds in an Effect RPC patch to handle large snapshots. The lockfile changes are mechanical hash updates from the patch modification.

^{You can customize Macroscope's approvability policy. Learn more.}

[t3dotgg]

I resolved the rebase locally, but this PR’s branch is in the organization-owned pearce-codes/t3code fork. GitHub rejects maintainer pushes to that fork, and the Update Branch API also reports a conflict. Please rebase the branch onto current main (the patch now targets patches/effect@4.0.0-beta.78.patch) or grant push access to the fork branch.

Dismissing prior approval to re-evaluate 6e69066

Dismissing prior approval to re-evaluate 644cc92

[Pearcekieser]

Thank for looking at this. Sorry for the delay was out on vacation. I recently started running into this timeouts again, so I'm hopeful this will help me and others.

I updated this to main. I don't think orgs owned branches allow pushes from upstream's maintainers, so I re opened this under a different PR if you need to edit.
New PR on a maintainer-pushable branch: #3817

Also @mincua suggested the same fix here so please try to give him credit if you merge this: #3500

Seems like this could also help fix:
#3553
#3734

Dismissing prior approval to re-evaluate 2825718

[juliusmarminge]

Closing because the author-designated maintainer-editable replacement is #3817. The replacement PR explicitly carries this work forward.

[Pearcekieser]

What Changed

Increases the patched Effect RPC pinger delay from 5 seconds to 60 seconds.

This changes the existing patches/effect@4.0.0-beta.73.patch so the generated RPC client waits longer before declaring the connection unhealthy due to a missed Pong.

Why

Large thread snapshots can occupy the WebSocket long enough that the Effect RPC Pong is queued behind snapshot chunks. On slow remote connections, the current 5 second heartbeat window can expire before the Pong reaches the client, causing a reconnect. The reconnect then requests the same large snapshot again, which can repeat the failure loop.

A 60 second window gives large snapshot sends more time to drain before the client treats the connection as dead.

This is a small mitigation for #2761. It does not solve the deeper issue that large snapshots and control/heartbeat messages share the same WebSocket queue; longer term, snapshots likely need pagination, truncation, backpressure, or control-message prioritization.

Checklist

• This PR is small and focused
• I explained what changed and why
• UI changes are not applicable

Note

Increase Effect RPC ping interval to 60 seconds for large snapshot support

• Updates patches/effect@4.0.0-beta.78.patch to change the protocol worker ping interval from 5 seconds to 60 seconds.
• Adds an hooks.onPing call (when provided) before each ping is sent, allowing pre-ping side effects.
• Behavioral Change: ping frequency drops from every 5s to every 60s, which may affect connection keepalive behavior in environments with short idle timeouts.

^{Macroscope summarized 2825718.}

---

Note

Medium Risk
Changes connection health detection for all Effect RPC WebSocket clients; a 60s window delays detection of truly dead peers but reduces false reconnect loops during heavy snapshot traffic.

Overview
Updates the vendored effect@4.0.0-beta.78 patch so the RPC socket pinger waits 60 seconds between ping attempts instead of 5 seconds, giving large WebSocket transfers (e.g. thread snapshots) more time before the client treats a missing Pong as a dead connection.

The same patch delta wires ConnectionHooks into the ping loop: onPong runs after recording a pong, onPing runs before writePing, and both default to Effect.void when unset. pnpm-lock.yaml is refreshed so every workspace consumer resolves the new effect patch hash.

^{Reviewed by Cursor Bugbot for commit 2825718. Bugbot is set up for automated code reviews on this repo. Configure here.}

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: 869b0f55-01e6-4fd4-8ceb-cad2894b8ac5

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[macroscopeapp]

Approvability

Verdict: Approved

Simple timeout configuration change from 5 to 60 seconds in an Effect RPC patch to handle large snapshots. The lockfile changes are mechanical hash updates from the patch modification.

^{You can customize Macroscope's approvability policy. Learn more.}

[t3dotgg]

I resolved the rebase locally, but this PR’s branch is in the organization-owned pearce-codes/t3code fork. GitHub rejects maintainer pushes to that fork, and the Update Branch API also reports a conflict. Please rebase the branch onto current main (the patch now targets patches/effect@4.0.0-beta.78.patch) or grant push access to the fork branch.

Dismissing prior approval to re-evaluate 6e69066

Dismissing prior approval to re-evaluate 644cc92

[Pearcekieser]

Thank for looking at this. Sorry for the delay was out on vacation. I recently started running into this timeouts again, so I'm hopeful this will help me and others.

I updated this to main. I don't think orgs owned branches allow pushes from upstream's maintainers, so I re opened this under a different PR if you need to edit.
New PR on a maintainer-pushable branch: #3817

Also @mincua suggested the same fix here so please try to give him credit if you merge this: #3500

Seems like this could also help fix:
#3553
#3734

Dismissing prior approval to re-evaluate 2825718

[juliusmarminge]

Closing because the author-designated maintainer-editable replacement is #3817. The replacement PR explicitly carries this work forward.