[lawrencecchen]

Summary

• E2-class fault found in the review of ghostty-next: the wrap edge break reads a compressed history row only when the pair was joined #29775. In ghostty-next, clearRows (ED 0, 1 and 2, DECALN) reset the cleared rows. But the row above the range kept wrap=true (a scrollback row for ED 2 and DECALN), and the row below kept wrap_continuation=true (the cursor row after ED 1). After a wider resize, reflow then joined the scrollback row and the new text into one line, for example "abcdeXY".
• Fix (fork PR terminal: breakWrapAbove reads a compressed history row only when the pair was joined ghostty-next#68, part 2): after a full clear, clearRows breaks the soft wrap across both edges of the range. A protected clear keeps its row flags. The breakWrapAbove comment that said "every clear clears both" is corrected.
• Pin: GhosttyNextKit xcframework-34e73ba8ba7cb8da0115caeb780065e3a12f2f62-apple-v6, on release/cmux-9ce9b7d0-e2. That release is 84a225487 (the ghostty-next: the wrap edge break reads a compressed history row only when the pair was joined #29775 pin) plus only this change. The gitlink, the notices and docs/ghostty-fork.md (release-branch commit IDs) move with it.

Testing

Red first. The focused command is cmux-ci run --class light --script scripts/ci/cmux-tui-rust-check.sh --arg test --arg breaks_the_wrap_from_scrollback --ref SHA.

• Red commit 62c4048 (old pin 84a225487): both tests in vt_clear_rows_wrap_edges.rs fail. Fleet step 0f6924a3ee8f463e83dca258 ran on the same commit on its first branch.
• Head (pin 34e73ba8b): pass. Fleet step 654683928964e92ce4f3cc76: 26 of 26 vt_ tests pass.
• Fork: red run 41251885912 (3 of 3 clearRows Zig tests fail without the fix). Green: fix cmux crashing issue #68 run 41251888717, broad run 41254020927 (2297 of 2297 libghostty-vt tests) and release run 41255186701, which published this kit.

Observed / Not run

• Observed: the runs above. The kit's SHA256SUMS lists 32cb09bbc69510a099df9206a460cc8e7f2aba41a1a5489de98f185da07154a4, which is the checksum in Package.swift.
• Not run: an app journey for ED or DECALN. The Rust tests check reflow after a resize on the same libghostty-vt.

Changelog

• Fixed: after the screen is cleared, text in the scrollback is no longer joined with new text when the window is resized.

[github-actions]

cmux status _{(one comment, edited in place; each line is updated by the job that owns it)}

Review efcaaf4: HQ feed card · gallery before/after runs · checks

App of this head (app.zip, kept 1 day)

gh workflow run reload-build.yml -R manaflow-ai/cmux --ref feat-cmux-next -f tag=pr30012 -f ref=cp/ghostty-clear-rows-wrap-edges-2 -f platform=macos -f backend_mode=local

Not verified: macOS jobs did not run: compile and app tests were skipped on efcaaf4ca8; ci-status is green without them.

CI passes on efcaaf4ca8 (run 41266146966 attempt 1).
macOS jobs did not run: compile and app tests were skipped: Fast static checks skipped (the macos job needs both to succeed). Re-run CI once they pass, before merging.

Written by scripts/ci/classify_failures.py (ci-failure-attribution.yml); signatures are its SIGNATURES table. A machine verdict is the runner's fault, not this PR's; yours means the failing file is one this PR changes, also red on main that main's latest full suite fails the same way, seen on other PRs that it failed on another pull request's run lately.

Shipped in:

• cmux NEXT nightly 1.0.0-nightly.4130052256501 (28aa638015)

[lawrencecchen]

Review (Opus 5.5 review subagent): SHIP at efcaaf4

Independent review subagent (read-only). It checked the clearRows edge breaks for every caller (ED 2 including the alt screen and after scrollClear, ED 0, ED 1, DECALN), the defer order and pin validity, the cases where bl is null or the last row and the protected clear, the cost (a compressed history page is restored only when row 0 truly continued it), the Zig and Rust test traces, and the URL and checksum (sha256sum -c of the release assets). It also checked that every doc ID is on release/cmux-9ce9b7d0-e2.
Minor, not blocking: no test covers ED 0 directly (eraseLine already resets that edge), but the Rust file's comment names ED 0.

Fleet evidence on this PR's exact commits:

• Red: step 62d8b159a972dfa5608ae475 on 62c4048. Both vt_clear_rows_wrap_edges tests fail.
• Green: step e1c74493eae7f9d131f10d0d on efcaaf4. 34 of 34 vt_ tests pass.

[blacksmith-sh]

Found 9 test failures on Blacksmith runners:

Failures

Test	View Logs
/tmp/tmp0ydl6b3_/summary-fail.json (live):/summary-fail.json (live):	View Logs
/tmp/tmp5nh_2hst/summary-missing.json (live):/summary-missing.json (live):	View Logs
[new_workspaces] debug.new_workspace timed 0 of 50 new workspace rows/ [new_workspaces] debug.new_workspace timed 0 of 50 new workspace rows	View Logs
[new_workspaces] the window draws no frames/[new_workspaces] the window draws no frames	View Logs
[tabs] debug.tab_switch timed 0 of 40 switches/ [tabs] debug.tab_switch timed 0 of 40 switches	View Logs
[tabs] the window draws no frames/[tabs] the window draws no frames	View Logs
convention violations found (ERROR lines above)./ convention violations found (ERROR lines above).	View Logs
test_every_part_of_bun_run_check_runs_in_the_warn_branch (main.WebviewsWarnSplitCov ersCheck.test_every_part_of_bun_run_check_runs_in_the_warn_branch)/ test_every_part_of_bun_run_check_runs_in_the_warn_branch (main.WebviewsWarnSplitCov ersCheck.test_every_part_of_bun_run_check_runs_in_the_warn_branch)	View Logs
test_no_runner_picker_returns_a_hosted_mac (main.NoHostedMacInNightlies.test_no_run ner_picker_returns_a_hosted_mac)/ test_no_runner_picker_returns_a_hosted_mac (main.NoHostedMacInNightlies.test_no_run ner_picker_returns_a_hosted_mac)	View Logs

^{Need help on this PR? Tag @codesmith-bot with what you need.}