Syncing Notes Across Devices

Nobody chooses a notes app for its sync. Then one day a paragraph is missing, or a file called note (conflicted copy 2026-07-14).md appears, and it becomes the only feature that matters.

Sync is where notes apps quietly differ most, and it is almost never in the comparison table. Here is what is actually happening underneath, and what to check before you trust something with years of writing.

The four architectures

1. First-party cloud sync. The app runs its own servers and the notes live there. Apple Notes, Notion, Evernote, Google Keep, Bear.

Nearly always the smoothest option, because the vendor controls both ends and can merge intelligently. The cost is that your notes live in their format on their infrastructure, and you should know what leaving looks like before you have four thousand notes. See export your notes.

2. File sync over a generic cloud drive. The app writes plain files; iCloud Drive, Dropbox, OneDrive or Google Drive moves them. Obsidian without Sync, Logseq, most Markdown apps.

The drive knows nothing about notes. It sees files, and it resolves conflicts by keeping both copies. That is safe and ugly. The classic failure is editing on a phone that has not finished downloading changes, which produces a fork rather than a merge.

3. Paid sync built for the app. Obsidian Sync, Standard Notes, Notesnook, Joplin Cloud. A sync service that understands the app's data, usually with end-to-end encryption and version history.

This is generally the right purchase if you are already on a file-based app and you use more than one device. It is the difference between "the vendor cannot read my notes" and "the vendor merges my notes properly", and these services mostly deliver both.

4. Peer to peer or self-hosted. Syncthing, a Git repository, a home server. No third party at all.

Genuinely private and genuinely more work. Syncthing needs devices to be awake at overlapping times, which is awkward with iOS in particular. Git gives perfect history and is unusable on a phone unless you enjoy resolving merge conflicts on a touchscreen. Both are excellent for a desktop-and-laptop setup and poor as a primary phone workflow.

The question that separates them

What happens when the same note is edited in two places before either has synced?

Every architecture answers this differently, and the answer is what you will live with.

• Merge at the character level. The best behaviour: both edits survive in one note. Real-time collaborative apps do this properly because they were built for simultaneous editing.
• Keep both versions. Safe, visible, annoying. Generic file sync does this. You lose nothing but you get duplicates to clean up, and people usually clean them up carelessly.
• Last write wins. The dangerous one. The device that syncs second overwrites the other, silently, and the earlier text is gone unless there is version history. Some lightweight and older apps still behave this way.

Before committing, run the test yourself. Turn off the network on both devices, edit the same note differently on each, reconnect, and see what you get. It takes four minutes and tells you more than any review.

Then check the second question: is there version history, and how far back? Conflict handling is prevention, and version history is the cure. An app with thirty days of per-note history is forgiving of almost every mistake, including your own.

Sync is not a backup

This is the misunderstanding that actually costs people their notes.

Sync propagates changes. A deletion is a change. Delete a folder on your laptop and a well-functioning sync service will delete it everywhere, correctly and fast. Corrupted file, same story. Sync replicates whatever happened, including the bad thing.

A backup is a copy that does not change when the original does. Snapshot it monthly, keep it somewhere else, and confirm you can actually open it. See how to back up your notes.

The related trap: if your only copy is in one vendor's cloud, your account is a single point of failure. Account lockouts are rare and not rare enough to bet a decade of notes on.

What to check before you commit

1. Conflict behaviour. Run the test above. Non-negotiable.
2. Version history. How far back, per note or per vault, and how easy to restore under stress.
3. Offline editing. Does the app work fully on a plane, and does it reconcile cleanly on landing? Some web-first apps degrade badly. See the best offline notes app.
4. Attachment handling. Images and PDFs are where file-based sync gets slow and where mobile clients quietly skip downloads.
5. Encryption. End-to-end means the provider cannot read your notes. Most mainstream apps encrypt in transit and at rest but hold the keys. Both are defensible; know which you have. See the best private notes app.
6. Scale. A few thousand small Markdown files sync differently from a few hundred notes with large attachments. Test with your real archive, not with ten notes.
7. What leaving looks like. Sync and export are the same question asked twice.

Practical rules that prevent most problems

• Let the app settle before editing. Most conflicts are created in the first ten seconds after opening an app that has been closed for a day. Open it, wait, then type.
• Capture on one device, edit on one device. Quick thoughts go into an inbox from anywhere; substantive editing happens in one place. This alone removes most conflict opportunities. See the note inbox.
• Prefer many small notes to few large ones. A conflict in a 200-word atomic note is trivial to resolve. A conflict in a 9,000-word master document is a bad afternoon.
• Do not stack two sync systems. A vault inside Dropbox that also uses the app's own sync is the most reliable way to corrupt something. Pick one.
• Keep the format portable. Plain text notes and Markdown can be repaired by hand when something goes wrong. Proprietary databases cannot.

The honest summary

If sync just working matters more than anything else, a first-party cloud app is the correct choice and the ecosystem lock-in is the price. If ownership matters more, use a file-based app and pay for its dedicated sync service rather than improvising with a generic drive. Improvised sync is where the horror stories come from, and it is nearly always someone saving ten dollars a month.

More in personal knowledge management, export your notes, and note-taking workflow.