How to choose a note-taking app you will still be using in five years

Every note app in this category can do headings, checkboxes, tags and search. Comparing those is how people end up switching tools every eight months, because the differences that actually decide whether a system survives are structural, and structural properties do not photograph well in a feature table.

There are four that matter, and they are worth more attention than the rest combined.

The failure mode of every note system is the note you did not write. If capturing a thought means choosing a notebook, a template and a title, you will not do it while you are in the middle of something else — which is exactly when the thoughts worth keeping arrive.

What good capture looks like: a keystroke that opens a search box which creates the note if nothing matches, a daily note that is always one command away, and a browser clipper for things you read. In Nodum that is ⌘O for the quick switcher, a daily note with your own template, and an MV3 Chrome clipper backed by a token that can only create notes.

One-directional links are what every app has: a hyperlink from A to B. The useful part is the reverse — standing on note B and seeing that A, F and Q all reference it, with the sentence around each reference.

This is the single feature that separates a note collection from a note system, and it is why the tools that have it — Obsidian, Logseq, Roam, Nodum — feel categorically different from the ones that do not. You get it for free at write time: type [[Compound interest]] and the backlink appears on the other side without any further work.

If your notes are rows in a hosted database, your relationship with the vendor is a tenancy. Pricing changes, feature removals and acquisitions are all things that happen to tenants. Markdown files in folders are the format with the longest demonstrated shelf life and the widest tool support, and they are readable in a text editor when everything else has gone.

This is not hypothetical: the largest waves of note-app migration in the last decade were triggered by pricing and ownership changes, not by better features appearing elsewhere.

Export is the feature you evaluate on day one and use on the worst day. Import fifty notes into a candidate, export them, and look at what comes out. Does the folder structure survive? Do the links still point at anything? Is it markdown, or an XML dialect that needs a converter?

Nodum's export is a folder-true zip of .md files — the same shape an Obsidian vault is, which is also what its importer accepts. That symmetry is deliberate: an import path that is not reversible is a trap, and a tool worth choosing should be one you can leave.

Against those four criteria: Obsidian wins on ownership and plugins but is proprietary and desktop-only. Notion wins on databases and team work but stores blocks, not files. Evernote and OneNote are strong capture tools with no linking model. Logseq is excellent if you outline. Nodum is the browser-based, open-source one — markdown files, wikilinks, automatic backlinks, a knowledge graph, and an export that gives back exactly what you put in.

Do this with any candidate before you commit a single real note to it.