Repository navigation
Replies: 1 comment
|
Most of this is already covered on main: late responses from a replaced search are discarded (first page, Load more, and the refill after a delete, including an account or folder change with the same text), a failed search clears the old rows and shows the error, and a "Searching…" status is announced while it runs. I'll keep previous results visible while a new search runs, since it feels faster, but dim them so they don't read as matches for the new query. The real gap is a failed Load more, which is silent today. I'll add an error next to the button so the retry is clear. Both go on main; I don't plan to adopt the clearing policy. |
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Is your feature request related to a problem? Please describe.
Could we agree on stricter result invalidation when replacing a search, including browser-history transitions? #592 is fixed, and the subsequent progress/geometry fixes in PRs #614/#622 are present in main. This proposal is about the remaining fork behavior, not reopening those reports or claiming their guards are absent.
Main intentionally keeps previous results visible during a replacement. The fork clears obsolete results immediately when the query/context changes, while keeping one reserved progress/retry slot. This makes it harder to mistake an old result for a match for the newly typed query. This is a UX choice for discussion, not a proven upstream defect.
Describe the solution you'd like
Acceptance scenarios with a delayed synthetic search endpoint:
Describe alternatives you've considered
Keeping previous results, as main does, can feel faster. An explicit previous-results label could make that equally clear without adopting the fork's clearing policy. Much of the sequence protection already exists upstream; only the residual contract and any missing edge tests need discussion.
Additional context
Checked main
8dbdc7982236404dbbb9de63dba769cfdb9ef701. Public fork commit, search/history/pagination tests, replacement search tests. Fork CI and synthetic Chromium/WebKit desktop/mobile checks passed. These are fork acceptance tests, not proof that every edge fails main. We would like to retire redundant fork behavior once upstream semantics are agreed.All reactions