Simon_Blair · 16 Oct 2025, 08:18 UTC
Stop automatic replay of that uncertain transfer and represent it separately from an untouched pocket; moving the save earlier would record intent, not prove the block arrived.
3 pointsAutomatic replay disabled. Yes, moving the save earlier was my first thought. Different gap, same problem.
18 pointsWhat actually confirms placement in your program? A release command returning isn't necessarily the destination receiving the block. The recovery screen needs the last confirmed physical step, not just the last line executed.
20 pointsOn a previous application, reopening recovery discarded the operator's unfinished checks, so include another restart while reconciliation is in progress once the main sequence is defined.
19 pointsChloe, agreed, though an unfinished operator check should remain unfinished after restart rather than become an authorisation merely because it was started.
12 pointsThomas, release completion only. Controls is defining destination confirmation. Chloe, recovery restart is now in the test list.
-8 pointsThanks for checking. Include source pocket, destination pocket and any potentially held block in the display. An index alone makes the operator guess which end of the transfer you're asking about.
7 pointsDan, yes, I meant preserve the uncertainty and supporting evidence, not restore a half-finished check as a completed decision.
12 pointsDraft now shows both ends and the uncertain transfer. No automatic recovery approved yet.
3 pointsHas the person covering the bench tried reading that draft without you explaining it? Useful next check before its labels harden into the handover.
5 pointsAlso test a genuinely completed transfer whose save succeeded; a repair that holds everything forever would avoid duplicate placement while still failing to recover useful work.
8 pointsThis discussion is closed to new replies after six months without activity. Last activity: .