Repository navigation
refactor(pty): delete relay-fed SSH output intake and projection threading - #27515
Merged
Merged
Conversation
Contributor
|
Important Review skippedReview was skipped as selected files did not have any reviewable changes. ⚙️ Run configuration
📥 Commits
You can disable this status message by setting the Use the checkbox below for a quick retry:
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info
📝 Walkthrough
Priority: ⬇️ Low Merge Risk: ⚪ Minimal · up to No merge-blocking issue is established; the gate commands do not add CI test runs. Pre-merge checks |
  |
| Check name | Status | Explanation | Resolution |
|---|---|---|---|
| Docstring Coverage |  |
Docstring coverage is 7.14% which is insufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 14 functions across 10 files. (1 skipped: … | Write docstrings for the functions missing them to satisfy the coverage threshold. |
✅ Passed checks (4 passed)
-
-
-
Check name Status Explanation Title check The title clearly and concisely describes the main change: removal of the relay-fed SSH output intake and projection threading. Description check The description is detailed and covers the change, rationale, linked-issue exception, visual proof, testing, compatibility impact, and AI disclosure sections. The Review and Checklist sections are not… Linked Issues check Check skipped because no linked issues were found for this pull request. Out of Scope Changes check Check skipped because no linked issues were found for this pull request.
-
-
Full details: Docstring Coverage
-
-
-
Explanation
Docstring coverage is 7.14% which is insufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 14 functions across 10 files. (1 skipped: 1 unsupported.)
-
-
✨ Finishing Touches 💡 1
-
-
-
📝 Generate docstrings 💡
-
-
-
- Commit to this branch
- Create a new PR
-
-
-
-
-
- Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts
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.
Comment @coderabbitai help to get the list of available commands.
OrcaWin
force-pushed
the
OrcaWin/p399-A3
branch
from
October 11, 2026 16:53
37be659 to
cdf372e
Compare
…ading Nothing in production feeds the SSH PTY output intake: its only data and exit entry points were called from tests, and the SSH PTY provider that would have fed it is gone. Delete the intake, its source-obligation ledger, model admission, projection publication and source-ack plumbing, and strip the projection ids and source-credit bookkeeping they threaded through the local renderer delivery path (accept, pending, flush, payload, accounting, exit, session). Local terminal delivery is unchanged. Because the intake also installed the runtime's remote source-range hooks, this desktop host now answers a terminal.subscribe that asks for ackOutputSourceRanges the way every orcad host already does: without the capability, so the client uses the byte-window ACK. No source ranges were ever attached to local output.
OrcaWin
force-pushed
the
OrcaWin/p399-A3
branch
from
October 11, 2026 17:10
cdf372e to
0522ebf
Compare
8 tasks done
OrcaWin
pushed a commit
that referenced
this pull request
Oct 11, 2026
The host side of the terminal source-range ACK protocol had one producer: the relay-fed SSH output intake, deleted in #27515. Since then nothing installs the runtime's source-range hooks and nothing passes source ranges into PTY output, so every terminal.multiplex subscribe that offers ackOutputSourceRanges is already answered without it and uses the byte-window ACK. Delete the ledger, the per-connection registry, the consumer hook surface on the runtime, the stream-level reservation/commit/rollback plumbing in flow control, initial snapshot and cleanup, the source-range ACK frame schema, and the range slicing in output chunks, batching and replay. The subscribe schema still accepts the ackOutputSourceRanges capability, and the client keeps its range-ACK support for older hosts. A new golden test pins that an offering subscribe gets the same reply and frames as a byte-ACK one.
OrcaWin
added a commit
that referenced
this pull request
Oct 11, 2026
The host side of the terminal source-range ACK protocol had one producer: the relay-fed SSH output intake, deleted in #27515. Since then nothing installs the runtime's source-range hooks and nothing passes source ranges into PTY output, so every terminal.multiplex subscribe that offers ackOutputSourceRanges is already answered without it and uses the byte-window ACK. Delete the ledger, the per-connection registry, the consumer hook surface on the runtime, the stream-level reservation/commit/rollback plumbing in flow control, initial snapshot and cleanup, the source-range ACK frame schema, and the range slicing in output chunks, batching and replay. The subscribe schema still accepts the ackOutputSourceRanges capability, and the client keeps its range-ACK support for older hosts. A new golden test pins that an offering subscribe gets the same reply and frames as a byte-ACK one. Co-authored-by: m4air <m4air@Mac.localdomain>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
ELI5
When Orca still drove SSH terminals through the old relay, the relay's output came into the app through a dedicated "intake". The intake tracked every chunk until the screen had shown it, so the relay could be told how much it was allowed to send next. The relay and the SSH terminal provider are gone (#27214, #27215, #27499), so nothing hands the intake any output now. It still starts with the app, and the local terminal path still carries its bookkeeping. This PR deletes the intake and that bookkeeping. Local terminals behave the same.
What Changed
src/main/ipc/ssh-pty-*.tsfiles: the intake and its registry, model admission, legacy projection, source-obligation ledger and coordinator, source-ACK coalescer and publication, generation guards, exit deadline and remote source-range consumers;ipc/pty/delivery/ssh-intake.ts, which built the intake when PTY handlers were registered;ipc/pty-pending-projection-admissions.ts;config/scripts/ssh-source-frontier-benchmark.mjs, a benchmark for one of the deleted files.projectionAdmissionIds/projectionAdmissionsTransferredfields are removed fromdelivery/{accept,pending,flush,payload,accounting,exit,wire-session,producer-sync}.ts,ipc/resize-visibility.ts,session.tsandpty-pending-data-drain-contract.ts. So aresession.sshOutputIntake,session.sourceCreditPendingPtys(only a relay span ever set it) and the intake cleanup hook inlistener-lifecycle.ts.sendPtyDataToRenderernow returnssentas a plain boolean. Every removed line was either asshOutputIntake?.…call on a field that was never fed, or a branch that ran only when a projection id was present. Local output never had one.pty-source-credit-contract.ts/-validation.tskeep only whatterminal-output-source-range.tsuses. Eleven exports that only the deleted files used are removed.pty-ipc-producer-flow-control.test.ts, one frompty-pending-output-cap.test.ts, and the intake-joinedit.eachinterminal-multiplex-source-range-admission.test.ts. Unusedvi.mocklines for the intake registry are removed from eight SSH IPC suites.reliability-gates.jsoncandci-shard-timings.jsondrop the deleted files; the gate checker passes.Net lines: prod +52 / −5,362 (−5,310), tests +2 / −3,725 (−3,723), config +7 / −366 (−359).
Why
Phase 3.99 (A3) removes code that is compiled, wired at startup and tested, but that nothing in production feeds. The projection bookkeeping sat in the terminal delivery hot path, so every change to local delivery had to keep it intact. The alternative was to stop installing the intake and leave the threading in place. That would leave dead branches on every flush, ACK, drop and exit.
Proof the deleted code is unreachable:
acceptSshPtyOutputDataandacceptSshPtyOutputExit, had no production caller. They were used only by two PTY IPC tests and the SSH IPC mock module. Its generation, migration and publisher entry points (allocateSshPtyProviderGeneration,beginSshPtyOutputGenerationMigration,installSshPty{SourceAck,SourceCancellation}Publisher,applySshPty*CancellationProof,getSshPtyAcceptedSourceCheckpoints,closeSshPtyOutputGeneration) were referenced only by that mock.ssh-intake.tsalready hard-coded "no producer" after refactor(pty): delete the SSH PTY provider and its registry slot #27499.rg -wfinds zero real hits insrc tests mobile config .github. The only matches are unrelated local names that happen to be spelled the same (retainedBytes,admissionError,SpanRecord).rg -wover all 159@ts-nocheckfiles finds zero hits.build:electron-vite,build:cliandbuild:orcadall succeed.One observable change, mixed-version safe (the design assumed this was dead, but it is not exactly): the intake was also the only thing that installed the runtime's remote source-range hooks (
setRemoteTerminalSourceRangeConsumerHooks). So a paired client that asks this desktop app forackOutputSourceRangesonterminal.subscribeused to get it granted. No source range was ever attached to local output, so those streams carried empty range lists. Now the desktop host declines the capability, which is what every orcad and headless host already does: the hooks arenullthere. The client keeps the byte-window ACK (ackOutputSourceRangesis parsed as=== 1and defaults to off). Every client version that requests the capability already handles hosts that decline it, so old and new clients both work. The request field stays accepted instream-schemas.ts. The host-side source-range ledger itself is deleted in A6, and that PR's golden test (a byte-identical reply) relies on this.Left for later PRs in the stack:
OrcaRuntime.acceptPtyDataBounded/resetPtyModelAfterMigrationFailure(A4), andsetRemoteTerminalSourceRangeConsumerHooksplus the source-range ledger (A6). They now have no production caller.SSH boundary / folder workspaces / Windows / WSL: nothing here decides liveness or routes work, so no SSH-owned state ever becomes local or exited. Folder workspaces, Windows and WSL paths are untouched; the WSL guest has its own transport.
Linked Issue
N/A (maintainer, Phase 3.99 cleanup plan, PR A3)
Visual Proof
N/A. No UI change. Local terminal output, flow control, hidden-pane drops and exit ordering are unchanged.
Testing
keeps legacy multiplex clients outside source replacement admissionand the orcad-host suites.src/main/ipc,src/main/runtime,src/shared,src/main/daemon(including the serialize-grid transcript replay),src/main/providers,src/main/startupandtests/e2e/cross-version-wire. 3,058 files and 33,130 tests passed.pnpm tc,oxlint,pnpm run check:code-quality:changed(0 new findings),pnpm lint(all ratchets OK),check-reliability-gates, andbuild:electron-vite/build:cli/build:orcad.AI Disclosure
Review
Agent skill upstream boundary