Distinguish an uncovered --real window from a missing symbol in the no-data refusal #242
Reference in New Issue
Block a user
Delete Branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Symptom (fieldtest cycle-241, 2026-07-12, engine at
72c0b8a)aura run <bp> --real GER40 --from <ms> --to <ms>with a window the archive does not cover (or an inverted window) refuses withalthough GER40 IS present in the archive (M1 coverage 2014-08 .. 2026-07) — the message misattributes an empty window to a missing symbol. A consumer reading it concludes the archive lacks the instrument and stops, when narrowing/correcting the window is the actual remedy. Reproduced inside a fresh data-only project with a wrong-year window; the same command with a covered 2024 window runs fine.
Expected: the empty-window case names the window (requested bounds + the symbol's actual coverage, or at least "no data in the requested window"), distinct from the genuinely-unknown-symbol case. Pre-existing behaviour, not a #241 regression (surfaced by the cycle's real-data fieldtest axis).