[lawrencecchen]

Summary

Reverts #30258 (06020a3), opened by gh-merge-green --revert.

Changelog

none

---

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

---

Summary by cubic

Reverts #30258, which let DEV builds update a fresh machine's older cmux-tui automatically when starting a new chat.

The machine-update-before-copy fallback in CloudStartChat is removed, and cmux-remote no longer resolves DEV-build pins from the commit's published manifest.

^{Written for commit fbd1824. Summary will update on new commits.}

Takes back #30258: CmuxNextApp +21 lines turned the feat tip arch ratchet red; re-land with the code moved to a package.

[lawrencecchen]

Review (subagent, correctness first): SHIP at fbd1824

• Exact inverse: git diff 06020a3aabc^ 06020a3aabc is byte-identical to git diff fbd18242263c11874210f9f395492987ed056fb4 fbd18242263c11874210f9f395492987ed056fb4^. The head tree equals 06020a3aabc^ for all three files. Base 951865e to head touches only CloudStartChat.swift, CloudStartChatMachineTests.swift and ssh_artifacts.rs.
• No later dependency: none of the 22 commits on feat-cmux-next after 06020a3 (tip f176022) touches these files. The tip uses no symbol that Cloud: a new chat on a fresh machine with an older cmux-tui updates it and goes on (DEV builds can update machines) #30258 added (updatesBeforeCopyingAgain, published_manifest, published_pin, is_commit, published_pin_tests). The only hits are unrelated Python names (verify_published_manifest.py, seen_this_commit). PUBLISHED_BASE_URL, fetch_published, PublishedPayload, release_target and ArtifactManifest existed before Cloud: a new chat on a fresh machine with an older cmux-tui updates it and goes on (DEV builds can update machines) #30258 and stay. The signature of pinned_linux_payload and its caller machine_update.rs are unchanged.
• app_lines (arch_counts.py, CmuxNextApp Sources only): CloudStartChat.swift goes from 189 to 168 lines, so -21. The -12 test lines are in the CmuxNextAppTests module, which the row does not count.
• The description has the take-back line ("Takes back Cloud: a new chat on a fresh machine with an older cmux-tui updates it and goes on (DEV builds can update machines) #30258: ...").
• Open PRs that build on Cloud: a new chat on a fresh machine with an older cmux-tui updates it and goes on (DEV builds can update machines) #30258: cmux link machine update: a tree-mode DEV build pins the build its cmux-tui tree published #30421 ("cmux link machine update: a tree-mode DEV build pins the build its cmux-tui tree published") edits published_manifest, published_pin and is_commit. After this revert it conflicts, and it must re-land on top of a re-land of Cloud: a new chat on a fresh machine with an older cmux-tui updates it and goes on (DEV builds can update machines) #30258. The draft HOLD Cloud start proof: integration of #29962 and #30162 (capture only) #30171 (capture only) carries the same code. Cloud: show startup checklist in chat #30386 and cmux NEXT Cloud: picking another machine for a Cloud chat whose first send failed moves it there and sends again #30002 touch CloudStartChat.swift but use none of the reverted code. Cloud: show startup checklist in chat #30386 reads copy.0, which works with the old let copy.
• Behavior cost (not a blocker): DEV builds again cannot update a Cloud machine image with an older cmux-tui. A new chat's first send on such a machine fails with "Update the machine" until Cloud: a new chat on a fresh machine with an older cmux-tui updates it and goes on (DEV builds can update machines) #30258 re-lands, with its code moved out of CmuxNextApp.
• Not run: no build or test on this machine (read-only review of the refs).

[github-actions]

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

Review fbd1824: 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=pr30438 -f ref=revert/pr-30258 -f platform=macos -f backend_mode=local

Shipped in:

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

[blacksmith-sh]

Found 12 test failures on Blacksmith runners:

Failures

Test	View Logs
/tmp/tmph3ywtora/summary-missing.json (live):/summary-missing.json (live):	View Logs
/tmp/tmpm1ir0yf5/summary-fail.json (live):/summary-fail.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
nightly must build and hand off the macOS 26 helper and Xcode 26.5 app before publishin g/ nightly must build and hand off the macOS 26 helper and Xcode 26.5 app before publishin g	View Logs
test_helper_never_takes_a_hosted_mac_and_skips_compile_on_hit (main.WorkflowTests.t est_helper_never_takes_a_hosted_mac_and_skips_compile_on_hit)/ test_helper_never_takes_a_hosted_mac_and_skips_compile_on_hit (main.WorkflowTests.t est_helper_never_takes_a_hosted_mac_and_skips_compile_on_hit)	View Logs
test_no_suite_runs_page_javascript_on_a_built_bundle (main.BundleInputs.test_no_sui te_runs_page_javascript_on_a_built_bundle)/ test_no_suite_runs_page_javascript_on_a_built_bundle (main.BundleInputs.test_no_sui te_runs_page_javascript_on_a_built_bundle)	View Logs
test_the_nightly_is_covered (main.NightlyStepCoverageTests.test_the_nightly_is_cove red)/ test_the_nightly_is_covered (main.NightlyStepCoverageTests.test_the_nightly_is_cove red)	View Logs

...and 2 more test failures. View all on Blacksmith

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