Day 0 decision
Pause the jar. Review the protocol.
The original sketch treated a sunny counter, a cloth-covered vessel, a salt estimate, and sensory checks as enough to begin. A safety review flagged that setup as unsuitable to publish as guidance. Those details are withdrawn rather than quietly polished into a recipe.
Fermentation safety depends on a complete, research-based process: ingredients, vessel, oxygen exposure, temperature control, acidity measurement, handling, storage, and discard criteria all interact. A visually convincing jar is not evidence that a process is safe.
The useful field note today is the stop: no batch begins until the process earns review.
Required before testing
What the safety review must settle.
- Authoritative starting pointIdentify a research-based, kanji-specific process or obtain qualified review of the full proposed protocol.
- Controlled environmentDefine the appropriate vessel, closure, oxygen conditions, temperature range, light exposure, and contamination controls.
- Measured formulationSpecify ingredients by weight and establish whether the full formulation—not an isolated salt percentage—has support.
- Verified safety checkpointsDefine what must be measured, how instruments are calibrated, which limits apply, and who reviews the readings.
- Stop and storage rulesSet unambiguous rejection, refrigeration, shelf-life, and disposal criteria before anyone tastes the batch.
Future field record
What a real log would document.
Only after approval, the Day 0 record should identify the reviewed source and protocol version, batch ID, ingredient lots and weights, equipment and cleaning record, instrument calibration, starting measurements, environment, named human operator, photographs, and any deviations.
Later entries should preserve measurements and observations without translating appearance, fizz, aroma, or taste into a safety verdict. Any change to the reviewed process should return the batch to review rather than being improvised in public.
Publication ladder
Three gates, in order.
- Qualified safety reviewA complete, kanji-specific protocol is reviewed and the source, scope, and unresolved limits are documented.
- Approved human testA named adult tester follows the reviewed protocol and records the batch without claiming success in advance.
- Editorial reviewThe log is checked against the record. Only then can observations be published, with any safety limits kept visible.