Skip to content
Log in

Bridges

Import & export

One-time files, in or out. Import people, connections or occasions from vCard, CSV, JSON or iCal, and export everything again whenever you like — to back up or to move elsewhere.

Your contacts are yours. Bring them in from wherever they live now, and take everything out again whenever you like.

Import & export is for files that cross once, when you ask: an address book you bring in, or a copy of your Space you take out. Open it from the Bridges hub.

A file never runs on its own: it moves when you move it, in either direction, and the formats are ordinary ones rather than an export only Longtime can read.

Importing a file

Start an import and pick the file. Longtime tells the format from the file's extension:

Format What it can hold
vCard (.vcf) People, with their contact details
Comma-separated values (.csv) People or connections — you say which, then review the columns
JSON (.json) People, connections or occasions — you say which, then review the fields
iCal (.ics) Connections, proposed for your review — all-day and multi-day events as occasions

vCard is the format worth reaching for first. A vCard (.vcf) is the contact-card file your phone or contacts app exports — Export in Google Contacts, for example, or File → Export → Export vCard in Contacts on a Mac. It carries names, phone numbers, email addresses, addresses and birthdays in fields Longtime already understands — so it needs no mapping at all.

Suggesting or applying

Next, choose how the records arrive:

  • Suggest — every record lands in your Suggestions inbox, for you to approve one by one.
  • Apply — records go straight in. Longtime adds details you do not have yet and never changes ones you do; where the file disagrees, it asks you in Suggestions.

Calendar files always arrive as suggestions, whichever you choose.

Give it a label if you like — "From my phone" — and start the import.

An import matches each record to the people you already have — by Longtime's own ID, email, phone or name — so it does not add a second copy of anyone it can recognize. If someone slips through, under a different spelling with no shared email or phone, Longtime suggests merging them. A file you exported from Longtime comes back to the same records.

Mapping columns

A CSV or JSON file goes to a mapping screen first, which shows each column in your file and where Longtime thinks it goes. Check each one, choose a field or skip the column, and confirm; Longtime remembers your answers, so the next file shaped the same way needs fewer.

Preparing a file

The surest template is one of your own: export a few people or connections as CSV, open the file, and use its columns. Longtime reads its own files back without a single question. For a file from somewhere else, these are what it looks for.

Column names. Longtime reads the header row, not the order of the columns, and knows the usual names for each field, in any capitals or punctuation:

For Columns it recognizes
Names First name (or Given name), Last name (Surname, Family name), Middle name, Name or Full name, Preferred name (Nickname, Known as), Previous last name (Maiden name)
Dates Birthday (Birth date, Date of birth, DOB), Death date
Contact details Email (Emails, E-mail address), Phone (Mobile, Cell, Telephone)
Addresses Street (Address line 1), Address line 2 (Apartment, Suite), City (Town), State (Region, Province, County), ZIP (Postal code, Postcode), Country
Everything else Circles (Groups, Labels, Tags), About, Time zone, Cadence count and Cadence type, and a pair for each dated note: Notes 1 Date and Notes 1 Text
Connections Date, Time, Start timezone, Description (Title, Subject), Type, Location, and for each guest Guest 1 and Guest 1 email, Guest 2 and Guest 2 email

A column it cannot place is marked on the mapping screen for you to choose, and a column you skip is left out.

Several values in one cell. Separate them with semicolons: ana@example.com; ana.work@example.com. Numbered columns work too, as most contacts apps write them: "Email 1", "Email 2", or "Address 1 City", "Address 2 City".

Dates. Any common form: 2026-03-02, 3/2/2026 or March 2, 2026. Longtime looks at each date column as a whole: if any date in it can only be read day first (25/12/2026; there is no month 25), the whole column is read day first, the way UK and European apps write dates. A column that never says is read month first, the US way, so when every date could go either way, year-month-day (2026-02-03) can't be misread. A birthday without a year is fine as 14 Mar. For a connection, a time can have its own column (18:30 or 6:30 PM), or sit in the date's cell: 2026-03-02 18:30. A row with a date and no time asks, as a suggestion, whether it is an occasion, which needs no time, or a connection at a time you give.

Addresses. One column per part reads best. An address missing its state, ZIP or country is kept as it came, never guessed at; when it arrives as a suggestion, the card says what is missing so you can add it.

People on a connection. One guest per numbered pair of columns, "Guest 1" and "Guest 1 email", then "Guest 2" and "Guest 2 email", as Longtime's own export writes them. (Several names in one cell can't be told apart, so give each its own column; a single guest can also go in a People column.) A guest found by email is certain; one found only by name is asked about: "Who is this?", with everyone it could be.

Notes and cadences. In a file, "About" is the About field and "Notes 1 Date" and "Notes 1 Text" are a dated note, one pair per note. (A single "Notes" column, as older files and other apps write it, goes to About.) Dated notes come back with the day they were written, and a note someone already has is not added twice. A cadence fills in only where the person has none; one you already set is never changed.

After an import

Each import in the list shows what happened — "4 added · 2 updated · 3 to review · 1 skipped". From there you can download the file you uploaded, run it again, or remove its record (which removes the import from this list, not the people it brought in).

If an import left suggestions waiting, removing it asks whether to keep them in your inbox or withdraw them too. People it already added stay either way.

Import adds; it does not replace

An import brings records into your Space. It never deletes anything, and it never overwrites what you have — so a file that is missing someone does not remove them.

Exporting

You can take everything out again, whenever you like. That is deliberate, and it is not a feature we would remove: a tool that holds your address book and cannot hand it back is a tool you cannot leave, and the point of paying month to month is that leaving stays possible.

Start an export, choose what to export — people, connections, occasions or contact info — and a format:

What Formats
People vCard, CSV, JSON
Connections iCal, CSV, JSON
Occasions JSON
Contact info CSV, JSON

For people, if you have Pools, you can export everyone in your Space or just one of them. Add an optional label — "Full backup" — and create the export. When the file is ready it appears in the list to download, and the Bridges hub says so too.

Contact info

Contact info turns the people you export around: one row per address, phone number or email address, with everyone who shares it on that row. It's what turns a Pool into a holiday-card run, a phone tree or a mailing list.

  • Which ones. By default each person appears once, at their preferred address (or number, or email); if they haven't marked one, at each of theirs. Choose Include every address to list all of them. People with none of that kind are left out, and the finished export says how many.
  • The same place, however it was saved. Copies of one address that differ only in capitals, spacing, punctuation or standard abbreviations ("Rd" and "Road") share a row.
  • Name each row for the household (everyone recorded there, or the household name you set on the address, such as "The Smith Family") or for just the people on the list ("Jane & John Smith"). Each Pool remembers your choice.

In CSV the people on a row are separated by semicolons; in JSON they are a list. Contact info is for using elsewhere, not for importing back.

Backing up

There is no separate "download my data" button, because this is it. Export people and connections as JSON or CSV, and occasions as JSON, and a fresh import rebuilds them, matched back to the same records. A JSON or CSV export of people carries everything about them: contact details (including each address's household name), circles, About text, time zone, date of death, cadence, and the dated notes on their page, each with the day it was written. Bring it back into the same Space and nothing is duplicated. A vCard carries contact details and time zone, for a contacts app, but not the rest. Pools are saved searches rather than records, so there is nothing of theirs to export; they take a moment to set up again.

The import and export records in these lists, and their files, are kept for 90 days, then deleted. The people and connections an import brought in are not affected. Download any file you need before then.

Details

Files can be up to 10 MB.

For CSV and JSON, Longtime asks what the file holds — people or connections for CSV, and occasions too for JSON. A vCard or a calendar says so for itself; a spreadsheet cannot. Of the files you name the contents of, only JSON can hold occasions.

With Apply, where the file disagrees with what you have, the difference goes to Suggestions instead. Records that might be someone you already have, when Longtime is not sure, go to Suggestions too. People beyond your Space's limit are skipped, and counted as skipped.

A calendar file follows the same rules as a calendar you bring in (see Calendar feeds) — only events with someone else on them, guests matched by email address — except for dates: a file's one-off events are proposed whatever their date, and a repeating meeting only for its occurrences in about the last three months. Uploading a calendar that Longtime exported, or the same file again, brings in only what changed.

On the mapping screen, Longtime says why it made each pick — "Named exactly in the file", "You've mapped this before", "Best guess from the values". Picks it is unsure of are marked "needs answer", and columns it could not place are marked "skipped". Choose a field for either, or skip the column.

A "Best guess from the values" pick comes from AI: for a column whose name does not say what it holds, the name and a few of its values are used to suggest where it belongs. It is only ever a suggestion — you confirm every column. See the AI notice.

Previous
Bridges