Page number option 2 is still showing even though it should not be showing as there is no adjacent page currently showing and its not a end page.
Page number option 2 is still showing even though it should not be showing as there is no adjacent page currently showing and its not a end page.
QA PASS (Fix attempt 1). Merged #239 (b840bff) into dev, deployed to palette-dev, verified live.
Evidence (rendered probe on live palette-dev, served table.js sha256 03e8b0d9... matches origin/dev):
30 pages, page 1: [1, 2, ..., 30] - page 2 shows only as current+1 (correct)
page 27: [1, ..., 26, 27, 28, ..., 30] - page 2 gone (old code hardcoded it here; this was the reported bug)
pages 25/26/29/30: window is 1, current+-1, last, no page 2
local build with 126 test pastes: unfiltered and filtered branches both verified, page 2 absent from non-adjacent windows, no ellipsis regression
console clean, no CSP violations
QA PASS (Fix attempt 1). Merged #239 (b840bff) into dev, deployed to palette-dev, verified live.
Evidence (rendered probe on live palette-dev, served table.js sha256 03e8b0d9... matches origin/dev):
- 30 pages, page 1: [1, 2, ..., 30] - page 2 shows only as current+1 (correct)
- page 27: [1, ..., 26, 27, 28, ..., 30] - page 2 gone (old code hardcoded it here; this was the reported bug)
- pages 25/26/29/30: window is 1, current+-1, last, no page 2
- local build with 126 test pastes: unfiltered and filtered branches both verified, page 2 absent from non-adjacent windows, no ellipsis regression
- console clean, no CSP violations
QA PASS (PR #239). Verified locally and live on palette-dev.
Local build from pr239 head (b840bff): go build + go test all pass; rendered CDP probe over a seeded 7-page history confirmed the orphan page 2 is gone at page 6 (pager: 1 ... 5 6 7 ...) while adjacent/edge cases stay correct (page 1: 1 2 ... 7, page 2: 1 2 3 ... 7, last: 1 ... 6 7).
PR was already merged into dev (merge commit 9490b8d, CI docker success). Rolled out to palette-dev; served /static/table.js sha256 matches git show origin/dev exactly.
Live check on palette-dev /history (736 pastes, 30 pages): page 1 shows 1 2 ... 30 (2 adjacent, correct); after paging to page 7 the pager shows 1 ... 6 [7] 8 ... 30 with no orphan 2, matching the issue screenshot scenario.
Closing and cleaning up.
QA PASS (PR #239). Verified locally and live on palette-dev.
- Local build from pr239 head (b840bff): go build + go test all pass; rendered CDP probe over a seeded 7-page history confirmed the orphan page 2 is gone at page 6 (pager: 1 ... 5 6 7 ...) while adjacent/edge cases stay correct (page 1: 1 2 ... 7, page 2: 1 2 3 ... 7, last: 1 ... 6 7).
- PR was already merged into dev (merge commit 9490b8d, CI docker success). Rolled out to palette-dev; served /static/table.js sha256 matches git show origin/dev exactly.
- Live check on palette-dev /history (736 pastes, 30 pages): page 1 shows 1 2 ... 30 (2 adjacent, correct); after paging to page 7 the pager shows 1 ... 6 [7] 8 ... 30 with no orphan 2, matching the issue screenshot scenario.
Closing and cleaning up.
Re-verified against the live dev build instead of re-fixing. The fix from #239 (b840bff, drop hardcoded page 2 from pagination window) is deployed on palette-dev: served /static/table.js sha256 03e8b0d9... matches origin/dev tip, and the old win = new Set([1, 2, ...]) is gone.
Rendered probe on live palette-dev (/history, 736 pastes = 30 pages), clicked through to page 6:
same result at 375x812 (mobile override) and after a reload with prefs reset
sorted/filtered path (the second window block in table.js) also verified: sort by title, last page 4 of 4 shows [‹] 1 … 3 4 [›], no stray 2
go build/test pass on a fresh fix-238-r2 clone of dev tip
The screenshot in this issue matches the pre-fix build (main still carries the old 1, 2, ... window and prod's table.js is the old one; prod currently has 10 pastes so the pager doesn't render there). Dev has been correct since the #239 deploy — closing as verified, no code change needed. Main will pick the fix up at the next release merge.
Re-verified against the live dev build instead of re-fixing. The fix from #239 (b840bff, `drop hardcoded page 2 from pagination window`) is deployed on palette-dev: served /static/table.js sha256 03e8b0d9... matches origin/dev tip, and the old `win = new Set([1, 2, ...])` is gone.
Rendered probe on live palette-dev (/history, 736 pastes = 30 pages), clicked through to page 6:
- pager renders: [‹] 1 … 5 6 7 … 30 [›] — page 2 absent, page 6 active
- same result at 375x812 (mobile override) and after a reload with prefs reset
- sorted/filtered path (the second window block in table.js) also verified: sort by title, last page 4 of 4 shows [‹] 1 … 3 4 [›], no stray 2
- go build/test pass on a fresh fix-238-r2 clone of dev tip
The screenshot in this issue matches the pre-fix build (main still carries the old `1, 2, ...` window and prod's table.js is the old one; prod currently has 10 pastes so the pager doesn't render there). Dev has been correct since the #239 deploy — closing as verified, no code change needed. Main will pick the fix up at the next release merge.
fen
removed the In Progress label 2026-09-10 22:57:44 +00:00
fen
added spent time 30 minutes2026-09-10 22:57:44 +00:00
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Page number option 2 is still showing even though it should not be showing as there is no adjacent page currently showing and its not a end page.
QA PASS (Fix attempt 1). Merged #239 (
b840bff) into dev, deployed to palette-dev, verified live.Evidence (rendered probe on live palette-dev, served table.js sha256 03e8b0d9... matches origin/dev):
QA PASS (PR #239). Verified locally and live on palette-dev.
b840bff): go build + go test all pass; rendered CDP probe over a seeded 7-page history confirmed the orphan page 2 is gone at page 6 (pager: 1 ... 5 6 7 ...) while adjacent/edge cases stay correct (page 1: 1 2 ... 7, page 2: 1 2 3 ... 7, last: 1 ... 6 7).9490b8d, CI docker success). Rolled out to palette-dev; served /static/table.js sha256 matches git show origin/dev exactly.Closing and cleaning up.
Re-verified against the live dev build instead of re-fixing. The fix from #239 (
b840bff,drop hardcoded page 2 from pagination window) is deployed on palette-dev: served /static/table.js sha256 03e8b0d9... matches origin/dev tip, and the oldwin = new Set([1, 2, ...])is gone.Rendered probe on live palette-dev (/history, 736 pastes = 30 pages), clicked through to page 6:
The screenshot in this issue matches the pre-fix build (main still carries the old
1, 2, ...window and prod's table.js is the old one; prod currently has 10 pastes so the pager doesn't render there). Dev has been correct since the #239 deploy — closing as verified, no code change needed. Main will pick the fix up at the next release merge.