Skip to content

test(linux): drop imported open times and fix two test races - #3500

Merged
datlechin merged 1 commit into
linuxfrom
fix/linux-red-after-lane-merges
Oct 18, 2026
Merged

datlechin merged 1 commit into
linuxfrom
fix/linux-red-after-lane-merges

Conversation

@datlechin

Copy link
Copy Markdown
Member

Why

linux CI has been red since #3467. Fast checks, Flatpak and every package job fail a_version_1_connection_list_imports_as_it_was_written. #3467 made an import drop a file's last-opened times, and #3464's fixture test still expected them. Since #3473, two app tests also time out on some jobs.

Change

  • The v1 import test expects no last-opened times. The fixture is unchanged.
  • an_unreadable_history_says_why_where_it_is_opened waits for win.show-history to turn on. The action follows the window's own history, which can still be opening, and activating a disabled action does nothing.
  • a_sorted_table_pages_by_offset serves a count of 1,000, like its sibling paging tests. With the old count of 2, a count landing after Next moved the pager back to the first page. That made a fourth read, and the wait for exactly three missed it.

Verified

stored_formats fails on linux and passes here. fmt, clippy and the tablepro lib suite pass in the Linux dev image, except sorting_10k_text_rows_finishes_within_one_second, which missed its 1 s bound at load average 56 and passes alone. Neither timeout reproduced locally; both races come from reading the code.

@datlechin
datlechin merged commit dd536e5 into linux Oct 18, 2026
15 of 16 checks passed
@datlechin
datlechin deleted the fix/linux-red-after-lane-merges branch October 18, 2026 02:23
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant