Corrections - picklerecipe.com

Reporting an error

The most useful report names the page, the value, and the document that says otherwise. This site does not hold opinions about fermentation; it holds transcriptions, so a correction is almost always either "the source says something different from what you printed" or "there is a better source and here it is". Both are welcome and the second is rarer and more valuable.

Every parameter on the site is printed with the sentence it came from, specifically so that a reader can tell those two cases apart before writing.

Corrections made

Date	What was wrong	What changed
2026-09-06	Every temperature taken from the FAO bulletin was ten times too high. The 1998 typesetting writes a degree sign as a superscript zero, so stripping the markup turned "21 to 24 C" into "210 to 240 C".	Caught before publication by the quote check. The converter now maps that markup to a degree sign before anything else runs.
2026-09-06	The organism parser split cells on punctuation, which produced entries like "Lactobacillus plantarum L. brevis L. fermentum L. fallax P. pentosaceus": one string, five organisms, and a taxonomy lookup that resolved to nothing.	Replaced with a matcher for the shape of a binomial, which finds all five. Organism count fell from 100 to 76, which is the correct number.
2026-09-06	The composition table matched a query word anywhere in a food's description, which put "Archway Home Style Cookies, Reduced Fat Ginger Snaps" on the ginger page.	The guard now matches only the primary name, the part before the first comma, which is what a FoodData Central description says the food is.
2026-09-06	Substrate pages were being created from Wikidata's "made from material" claims, which include the salt, the water, the vinegar and the oil. There was very nearly a page called /substrate/water.	Non-substrates and bare Latin binomials are now rejected from the substrate index; the botanical name has its own field.

Known limits

Degrees salometer is not converted. Four FAO brine figures are on that scale and the definition of the scale could not be sourced to a citable authority. They are printed as published. If you can point at a definition in a government or standards document, that is the single most useful correction anyone could send.
10 organism names do not resolve. Most are typographical errors in the source papers - confuses for confusus, germentum for fermentum. They are printed as published rather than corrected, because deciding which organism an author meant is a judgement this site is not entitled to make.
Region names are inconsistent because the sources are. One paper says "Korea", another says "East Asia", a third says "India, Nepal". Only obvious aliases are normalised; the rest are left as written.
The index is anglophone. It comes from English-language sources, so a pickle with no English-language literature is not in it. That is a real and known bias, and it is likeliest to be hiding Central Asian, West African and Latin American traditions.
Two romanisations may still be two pages. Three known pairs are merged by alias. Others almost certainly remain.

Things that look like errors and are not

Blank parameter fields. Most of the table is blank. That is the state of the record, not a loading failure. See methodology.
A brine figure of 15 to 20 percent. That is a commercial cucumber brine-stock process, not a household kraut. The parameter is printed next to the tradition it belongs to for exactly this reason.
Two different salt percentages for the same food. Where two sources disagree, both are shown with their sources rather than one being chosen.
An organism under an old name. The source's name is what is printed in the source's own row. The current name is on the organism page.