What Not To Can

Most of this subject is a list. Almost nobody who publishes the list explains it, and the explanation is the part that matters, because "not recommended" turns out to mean three completely different things depending on which item you are looking at.

Reason one: tested, and the answer was no. Somebody ran the work, characterised a hazard, and published a refusal. Pumpkin butter is the clearest case — home-made batches were measured as high as pH 5.4, which is low-acid, and the thickness and water content vary so much between batches that no single process can be calculated for them [F-9.3].

Reason two: never tested at all. No laboratory ever ran the thermal process work, so there is no time and no pressure to give you. Butter is the example, and the people who would have run that test say plainly that they do not know what the risk is [F-9.2]. That is an absence of research. It is not a finding of danger, and a site that reports it as one is lying to you in the cautious direction, which is still lying.

Reason three: tested once, then withdrawn. A published process existed and was pulled. Shelled nutmeats in a dry pack is that case: a University of Georgia process was in circulation, newer work on Salmonella in low-moisture foods undercut the assumption behind it, and there is no replacement [F-9.8] [F-14.10]. This is the shape nobody writes about, and it is the most honest thing in the whole subject — tested guidance is a moving object, and sometimes it moves backwards.

Keeping those three apart is the point of this hub. If you cannot tell which one you are looking at, you cannot tell whether to expect the guidance to change, and you cannot tell how strong the "no" actually is.

The six pages. The short list is the spine: every food on it, sorted by which of the three reasons put it there. Butter, milk and cheese takes the never-tested case in detail, because dairy is where the folk methods are thickest and where the tested source is at its most usefully honest about what it does not know. Pumpkin butter, purées and thickness is the physics page: why a dense food is a safety problem rather than a texture one, and what happens when a widely-shared recipe asks a boiling pot to reach a temperature it cannot. Bread and cake in jars is short, and includes the good news that some of the pages teaching it have been taken down or rewritten by their own authors. Green beans in a boiling water bath is the long one, because a recipe portal still hosts a plain-green-bean water-bath recipe with a popularity badge on it, and taking that apart properly is worth more than another list. Open kettle, oven, inversion, dishwasher, sun does the methods rather than the foods: one mechanism each, one citation each.

Every page here ends in a "do this instead." That is a rule, not a courtesy. A page that stops at "no" hands you straight back to the page that says yes, and that page is usually better written. Freezing, refrigeration, drying, or the tested version of the same food in a different form — there is almost always somewhere for the food to go.

And nothing in this hub is selling you anything. No book box appears on any of these pages, by design. A page whose whole content is "do not do this" with an advertisement underneath it is not a page you should trust, including when it is ours.

If you already did one of these things, the answer is not on this hub and it is not complicated: do not taste it to check. What a jar looks like tells you very little, because the organism at issue produces no reliable smell or appearance, and a lid that pings proves a vacuum formed rather than that anything was heated [F-1.9].