The Storm Drill Log That Could Make Code Less Guessy
The town hall siren wails, and volunteers jog their storm drill. Then it slips: the key ring is on the wrong hook, a radio dies, and people start asking, "Now what?" Reading the drill is like reading code, but running it shows what really happens under pressure.
A lot of code-writing tools mostly learn from what code looks like on the page. That helps them write tidy, familiar patterns. But it can miss what the code does when it runs, like slowdowns, memory strain, or weird corner cases. Like judging the drill by the binder, not the sweaty run-through.
So you try to record more run-throughs. Then reality bites: some drills cannot be run because something is missing, and notes are messy or copied. One team writes down times, another writes down feelings, so comparing runs turns into arguing. Writing down everything can even slow the drill itself.
The proposed fix is a shared log sheet that matches real action. Each run gets written as a simple back-and-forth: what someone did and what happened next, plus quick checkpoints like when a door is confirmed locked. That kind of run log is to code what the drill sheet is to the siren day, it captures behavior you cannot trust your eyes to spot in the written plan.
Then the proposal stacks those sheets in a careful way. One organized collection holds many run logs for the same kind of drill, even when teams wrote different versions of the plan, and it can keep practical notes like how long a run took or how much of the plan got used. A bigger stack holds repeats across conditions, like different buildings or shaky radios.
With a growing, open library like that, future code-writing tools could be checked against what actually happened in past run logs, not just what sounds right in text. Back in the town hall, the binder still matters, but nobody trusts it alone. The next siren can feel less like guessing and more like proof.