Decisions
112 recorded. None closed. So we stopped pretending.
A decision is not a table here any more. It is a knowledge entry with
type: decision — the same place SOPs and
references live, because that is what a decision turned out to be.
Six months, and not one was ever closed.
112
Decisions recorded
0
Outcomes ever entered
Recording one was easy, because it rode on a sentence somebody was already saying. Closing one meant remembering a row existed and going back to say something about it for its own sake. Nothing in a conversation ever prompts that, so it never happened — not once, in six months of daily use.
The tempting reading is that this is a discipline problem and the answer is a reminder. The honest reading is that a concept nobody ever completed was not carrying its own weight, and the fix was to stop asking.
All 112 moved. Nothing real was lost.
- Title, choice, context, reasoningCarried across into the entry’s title and body, in that order.
- The initiative it was aboutAll 65 links preserved, so a decision is still findable by the work it concerned.
- When it was decidedKept as the entry’s date, so the chronology survives the move.
- Outcome, review date, notes, confidenceDropped. Every one was empty, or read “pending” in all 112 rows.
Those four columns were counted before the migration was written rather than after it ran. Had any held content, this page would have to say what was given up.
The same question, one fewer place to look.
Ask what was decided about something and it searches one knowledge base rather than choosing between a decision log and a pile of notes. A decision that no longer holds is marked outdated, exactly as a superseded SOP is. Nothing is deleted.
You ask
What did we decide about broad match on Removery?
One search. Decisions, SOPs and references all answer.
It is a switch, not a demolition.
The original table still holds all 112 rows, read by nothing. Turning the decisions module back on restores the original tools reading the original records, outcomes and all. The fold is a judgement about how somebody wants to work, so being wrong about it costs one instruction rather than a restore.