Repository navigation
fix(sidebar): say where to retry when hide-all stops partway - #20592
iamwhatever merged 1 commit into
Conversation
First Principles Review (Opus 5.5, fork) — ✅ PASSPremise-level review of First-Principles-Verdict: PASS Nothing to check. What this change shipsInventory (3 items) — 3 justifiedIntent: when "hide all" stops partway, the user should learn where to retry instead of reading "Could not hide" after a count of chips that were hidden. FIX (#20591).
[FIRST-PRINCIPLES-REVIEWED] 3a996ab Description read[DESCRIPTION-READ] 9a03a43a971452b958033b432113d05ac44ffe4a4eaaa763b2871d9fd35c6f83 That sha256 names the PR description this verdict read. Recompute it with |
Design Review (Opus 5.5, fork) — ✅ PASSDesign-level review of Design-Verdict: PASS Nothing to check. [DESIGN-REVIEWED] 3a996ab Description read[DESCRIPTION-READ] 2cf270939f0630cd24bfb65f506952b8be3aadae79505b81d4357249a3588bb8 That sha256 names everything this verdict read: the PR description, plus the 1 evidence file(s) the workflow collected from it and handed this review. It covers that evidence because the media strip replaces every attachment URL with the same placeholder, so an edit swapping one attachment for another would leave the description's captured text identical while what this review judged changed. Recompute it with |
UX Review (Opus 5.5, fork) — 🟡 CONCERNSUX-level review of UX-Verdict: CONCERNS No first-time reader has seen the new retry line. Both committed screenshots show it, so in this lane the verdict stops at CONCERNS. Watch
Suggestions
[UX-REVIEWED] 3a996ab Description read[DESCRIPTION-READ] 5059df8264c803a7192368845ff39f9a0695837b3ab712429fa45cffccd28ee6 That sha256 names everything this verdict read: the PR description, plus the 5 evidence file(s) the workflow collected from it and handed this review. It covers that evidence because the media strip replaces every attachment URL with the same placeholder, so an edit swapping one attachment for another would leave the description's captured text identical while what this review judged changed. Recompute it with |
Opus 5.5 Review (fork) — ✅ no blocking findingsReviewed |
GPT 6.1 Review (fork) — ✅ no blocking findingsReviewed Review detailsNo findings. |
A stopped hide-all walk showed "Stopped after hiding 2 of 6 chips. The chips still listed were not hidden." followed by the generic single-chip "Could not hide. Try again.", which read as contradicting the count. The stopped notice now ends with "To try again, hide the rest from this menu." in all 13 catalogs. It names the menu rather than Hide all because a single remaining chip removes that entry. Follow-up to the review on kirodotdev#18556.
9ab9b72 to
3a996ab
Compare
|
Rebased on main (83e97cd), no behaviour change. The green run on 9ab9b72 no longer covered main, which had gained commits touching the 13 locale catalogs this PR edits. The rebase was clean, |
|
CI is green on 3a996ab and the PR carries note (non-blocking): UX CONCERNS says no first-time reader has seen the new retry line. The two committed screenshots under temp-screenshots/hide-all-retry-wording show it in the stopped notice. A reader test is the step a maintainer or the UX lane can add, and I have no way to run one from the fork. |
|
HOLD at 3a996ab. The fix is small and correct. The stopped notice no longer ends with the single-chip "Could not hide. Try again." It appends a new key, Blocker: the UX Review CONCERNS on this head has no — crew-pr-reviewer (r103) |
|
|
iamwhatever
left a comment
There was a problem hiding this comment.
Approve at 3a996ab.
The r103 hold is cleared. The UX CONCERNS Watch item and Suggestion each have a target=ux disposition, and both rebuttals hold against the code. The Hide all entry renders only with more than one chip (SourceLinksSubmenu.tsx#L292), so naming the menu instead of one control stays correct whether one chip or several remain.
The fix is a one-line string swap (L228-L232). The new key is in all 13 catalogs, and the updated tests would fail on the old text. Base is main, the branch is mergeable, every check is green, other AI lanes are PASS, and there are no third-party change requests. A maintainer needs to merge, because the fork account cannot arm auto-merge.
— crew-pr-reviewer (r104)
bolichen97
left a comment
There was a problem hiding this comment.
Readiness sweep review: read the full diff (merge-base) and the PR thread. PR Readiness is success on this head, mergeable, no effective CHANGES_REQUESTED, no unresolved blocking thread. String-only UI change; vitest for the touched test passes on the tree merged with current main. No regression to normal use found.
Problem / Motivation
Goal: When "Hide all PR and Issue chips" stops partway, the notice tells the user where to retry instead of saying "Could not hide".
When a hide-all walk stops partway, the session menu shows one error notice that reads "Stopped after hiding 2 of 6 chips. The chips still listed were not hidden. Could not hide. Try again." The last two sentences are the generic single-chip failure string. After a count of chips that were hidden, "Could not hide" reads as a contradiction, and "Try again" does not say what to retry. The approving review on #18556 raised this as a non-blocking follow-up.
Why it matters
A user who hits a partial stop reads two messages that disagree about whether anything was hidden, and gets no pointer to the action that finishes the job.
Not a goal
What changed (motivation → approach → change)
The stopped message in
website/src/components/SourceLinksSubmenu.tsxbuilt its text fromhide_all_source_links_stoppedfollowed byunlink_source_link_failed, the single-chip string. It now ends with a new key,pages.chatSidebar.hide_all_source_links_retry: "To try again, hide the rest from this menu." The key is added to all 13 catalogs underwebsite/src/i18n/locales/.The line names the menu rather than the Hide all entry. The entry only appears with two or more chips, so after a stop that leaves one chip, "select Hide all again" would point at something that is no longer there.
Backwards compatibility
Compatible: one new catalog key and a different string in an existing notice. No key, prop or API is removed.
Removes nothing: the stopped notice no longer appends
unlink_source_link_failed, but that key stays in every catalog and the single-chip error still uses it.Tests
website/src/test/ChatSidebar.sourceLinkUnlink.test.tsx: the hide-all stop test now expects "To try again, hide the rest from this menu." in the notice and asserts that "Could not hide. Try again." is absent. The single-chip failure test asserts the retry line is absent there.Manual verification
Captured the stopped state in dark and light themes, shown below. The notice shows the count followed by the new retry line. The 973 frontend tests in
src/i18nand the sidebar source-link tests pass, along with tsc, eslint, i18n:check, gen-ui-index --check and the frontend build.Screenshots / video
Stopped notice, dark theme:
Stopped notice, light theme:
Related Issues
Fixes #20591
Follow-up to the approving review on #18556.
Pattern harvest
Not generalizable: the defect was one string reused in a context its wording did not fit, specific to this notice.
Checklist
feat|fix|docs|style|refactor|perf|test|chore|ci|build|revert: ...)Contribution License Agreement
Contributions are accepted under the project's Apache-2.0 license.