Repository navigation
Show a Timeline commit's own diff on the Code tab - #71
Conversation
Where the host gives a diff, each commit's rows on the Timeline are a control. Enter or a click picks that commit on the Code tab and shows it, as the desktop's View commit button does. The same commit keeps the Code's scroll, and another starts it at the top with the list off the diff closed. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…anch Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…7e50cd Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
… again after a theme change Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
… page recipe Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…r in the Code recipe Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The Code recipe clicks a Timeline commit and checks the read of its diff, comes back to it by key and finds the Code as it was with nothing read, opens the other commit at its top, and finds the rows plain where the host gives no diff. The page file names the new status text. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
… past u64::MAX and reads the Code's diff again each time it shows a minute after it came Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…043b63, the environment switcher Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…ine count at the highest line number Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…act in the boundary unfold test Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
PASS for head
Regression sweep: 13/13 own recipes and master's 12/12 recipes pass on both builds. The base results are reused from the independently driven #70 head under Rules: all completed runs pass Tests: local CI: all four checks completed successfully on this exact head, including macOS and E2E smoke. Greptile: two resolved threads; no unanswered finding. Both relevant feature files are updated. Compared the cited Timeline button and openCommit behavior at f570bd2. Suppressing the unsupported-host action is the documented departure. Binary identities, LC_UUID from
Both build logs confirm Notes: no product finding. No pull request or branch was changed. |
… lays review threads out again when the theme changes Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…comments out again when the theme changes Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
PASS+NOTES for head
Regression sweep: all 13 checkout recipes pass on 4eb3fbe and 7dccbaf after the timed refresh rerun. Every current-master recipe was attempted on both binaries; 11 of 13 pass after that rerun, with the two shared baseline limits below. Every unchanged feature has matching settled Rules: doctor passed. Successful drives and recipes passed Evidence: Binaries driven, recorded LC_UUID values:
All newly built binaries showed Notes, shared baseline limits:
The first timed refresh attempts crossed a wall-clock gap from 13:34 UTC to 17:10 UTC and timed out on all builds. Fresh runs with idle sleep prevented pass the unchanged recipe. The original failures and successful reruns are retained. Screens differ only in fixture ports, ages, elapsed times, spinner frames and intermediate network/read states outside the intended change. Settled Summary and reviewer captures at 200x69 and 80x24 match, and settled requests match. |
Where the host gives a diff, a commit on the Timeline opens its own diff on the Code tab, as the desktop's
View commitbutton does.Enter view commit 9dc565a, and Enter or a click shows the Code with that commit picked in its list of commits.Stacked on #70.
All three come from the fixture server with #7's invented commits at 120x40.
Decision log
What I noticed. #70 added the Code tab's list of commits, and its docs left one known limit: a commit on the Timeline didn't open its diff, where the desktop's does.
How I confirmed it. In the pinned T3 source (f570bd2):
CommitEventas a<button aria-label="View commit {id}">that callsonOpen(event.id), and 599 passes itonOpenCommit.openCommit(oid)pick the commit for the Code tab and show that tab, and 2778 hands it to the Timeline.capabilities.diffis false and moves a reader on it to the Summary. The commit stays a button there, so a click lands on the Summary.Options.
Controlhold aStringand loseCopy.view commitfor the status line, from the desktop's label. The list of commits saysshow commit, which keeps its own wording.What review changed.
view commit 9dc565a.d25e836 and bdf01da merge #70 at b7d2adc and cebb4a4, which bring #66's fix for a hunk that removes nothing at the highest line number (8073497) and its test, with the same patch.
11720bf and 7dccbaf merge #70 at f14b610 and 4eb3fbe, which bring the fixes that lay a thread's comments and its loaded pages out again when the theme changes (#68's ad1590c and #69's 942dad5).
Known limits.
Tests
🤖 Generated with Claude Code


Confidence Score: 5/5
This PR appears safe to merge; no actionable issues remain.Summary
Timeline commits now open their own diff on the Code tab when the host supports diffs.
No new actionable issues found. Tests and terminal captures were not run during this review.
Diagram
%%{init: {'theme': 'neutral'}}%% flowchart TD A[Choose a Timeline commit] --> B{Host offers a diff?} B -->|No| C[Rows stay plain text] B -->|Yes| D{Same commit?} D -->|Yes| E[Restore Code reading position] D -->|No| F[Reset Code and request commit diff] E --> G[Show Code tab] F --> GReviews (9) · Last reviewed commit: "Merge the commit-scope branch at 4eb3fbe..." · Reviewed by Greptile