Undo a survey you should not have filed
Whose job: the engineer. You should see the run is retracted and stays on the record; condition falls back.
What you are doing
You are retracting a survey that should not have been filed — the wrong road, the wrong file, the wrong month.
This is the step people need under pressure, when something has gone in wrong and somebody senior is asking. It is listed with the sequence rather than hidden in a reference section for exactly that reason.
It is deliberately not the surveyor's. Entering what you saw and deciding a filing should be withdrawn are different acts, and the second is a decision about the council's record.
What PaveKeep does with it
The retraction, on the run itself: who filed it, who took it back, and when.
Nothing is deleted. The run stays on the record and drops out of every current figure; its segments fall back to whatever the previous survey said, or to "not surveyed" if this was their first. An auditor asking what happened to those numbers gets an answer, which is the entire difference between retracting and deleting.
What it will not do
- The method — Retracting a survey does not undo what was decided from it. A programme already run, a document already sent, a job already booked — those happened, and the record shows the condition they were based on and the fact it was later withdrawn.
- The product — A void is not reversible here. It is the recovery, and giving the recovery its own recovery is where these mechanisms stop being comprehensible. If a run was retracted in error, file it again.
- Your role — Retracting needs its own capability, which the surveyor role deliberately does not carry. If you cannot see the button, the answer is that this is somebody else's decision rather than a missing feature.
- Your subscription — The survey ledger is part of condition history. Without that module a council can still import condition; what it loses is the dated record of runs, and with it this way back.
How you know it worked
Before: a filed survey you should not have filed.
Do: Import Data, Surveys, void the run.
After: the run marked retracted with your name against it, still listed; the affected segments showing their previous condition or "not surveyed"; and every derived figure moving with them.
If you retracted the wrong one, file it again. The retraction stays on the record, which is honest — it is what happened.
Hand to: whoever received the figures that came from it. The product corrects itself; it cannot recall an email.
If it went in wrong: there is nothing to undo at this step. What to do instead is under what it will not do, above.