Why this is a different question from "which app is best"

Most software is judged on what it does this week. A family collection has to be judged on what happens to it in the years when nobody is thinking about it at all — the year you skip a renewal, the year the developer stops answering, the year the format changes, the year somebody dies and the phone is locked.

Recipes are the opposite kind of object from most files: they matter more every year they survive. Almost everything that goes wrong with a family collection goes wrong on that timescale, and none of it shows up in a review.

1 — Export, and check it before you trust anything

This is the whole checklist compressed into one item. Can you get everything out — recipes, notes, photographs of the cards — in a format that opens with no software from the vendor involved anywhere in the process?

Do not take the feature list's word for it. Run the export on day one, on ten recipes, and open the result in something ordinary. An export that arrives as a proprietary blob, drops the images, or lives behind a higher tier is not an export; it is a promise about an export.

2 — Does it hold the originals, or only the transcription?

A tool that stores only typed text has quietly decided that the handwriting was not the valuable part. For a household collection that is fine. For a family cookbook it is the whole loss.

Look for a photograph attached to the recipe rather than a photo album stored somewhere alongside it, because a link between two apps is exactly the thing that does not survive a decade.

3 — Can it produce something printed?

Paper is the only format in this list with a demonstrated forty-year track record, and a bound book is the only version of a collection that does not need a device, an account, a battery, or a company.

A tool that can set what you have typed into something a printer will accept has given you an exit that no technology change can close. A tool that cannot has made your collection permanently conditional on screens.

4 — What does it require of the people you share with?

There is a real difference between giving your sister a copy and giving her an invitation. A copy is hers, forever, with nothing to renew. An invitation is a relationship with a company she did not choose, and it lasts as long as that company and that account do.

Ask what your least technical relative has to do to read a recipe. If the answer involves creating an account, that is a cost the collection pays every time somebody new wants in.

5 — Does it work with nothing switched on?

Offline is not a convenience feature here; it is the test of where the data actually lives. If the recipes are readable in a basement kitchen with no signal, they are on the device. If they are not, they are somewhere else, and somewhere else is a place that can change its terms.

The same test answers the harder question underneath it: what happens if the company disappears. A collection that reads offline and exports whole has already answered it. Everything else is reassurance.

6 — What shape is the payment, and what does it imply?

This is last on purpose, because it is the item people argue about first and it matters least on its own. A one-time purchase is not automatically safer than a subscription — a bought app can be abandoned just as thoroughly as a rented one.

What the shape tells you is what the maker had to build. A program that must keep collecting has to keep you inside it; a program bought outright has no reason to make leaving difficult, which is usually why its export is better. Judge the export, and treat the pricing as evidence about the export rather than as the answer.

Where Heirloom lands on its own checklist

Applied to us, since a checklist you exempt yourself from is marketing: recipes stay yours, on your device, so it reads with nothing switched on. The card photographs are attached to the recipes, not stored beside them. The plain export and the print-ready book are part of the program rather than a tier above it. Sharing is a copy, not an invitation. And it is one payment of $29.99 — a keepsake, not a service.

The point of the list is not that we score well on it. It is that you should run it, on us and on everything else, before an afternoon of typing goes anywhere.