Repository navigation
Moving From An Earlier Version
This page is generated from
docs/LEGACY_EXPORT_PROCEDURE.md. Edit the source in the repository and the change will be synced here.
ReboundOps does not read another application's files or folders. To bring data across, you produce one JSON file in the neutral import format, check it, and then import it. Every step is read-only for the earlier version.
This procedure has not been run on real data yet. Its steps are tested only on synthetic files.
- Do this in a separate session that you authorise for the purpose. That session needs read access to the earlier version's data. ReboundOps' own development sessions are not given that access.
- Take a ReboundOps backup first (Settings → Backup and restore; a password-protected backup is recommended because it holds personal data), so you can return to where you were.
- Work in a new staging folder outside both projects, for example
~/reboundops-migration/. Do not put it in a Git repository. It will hold personal data.
- Find where the earlier version stores its data (its own documentation or settings will say).
- Copy that data into the staging folder. Use a plain copy, such as
cp -Ror the Finder. Do not open the original with tools that might write to it, such as a database browser in read-write mode. - Check that the copy is complete (file count and sizes) and then work only on the copy.
Write a small one-off converter (in any language) that reads the copy and
writes reboundops-import.json in the format described in
IMPORT_FORMAT.md. Rules:
- Map only what the earlier data actually contains. Leave a field out rather than guessing. Do not invent dates, employers, salaries or statuses.
- Use
status: "applied"(or later) only where the earlier data recorded that you submitted the application. IncludeappliedAtwhen the date is known. - Put free-text notes in
notes, and status changes inhistory. - Never include passwords, API keys or tokens.
npm run import:check -- ~/reboundops-migration/reboundops-import.jsonThis prints what would be added, which records would be rejected and why, and any warnings. Fix the converter and repeat until there are no unexpected rejections. Nothing is written by this command.
- Open ReboundOps and go to Settings → Import data.
- Choose the file. The preview shows what will be added, which jobs already exist (they are skipped), and which records were rejected. Nothing is saved yet.
- Confirm to import. The import runs in a single transaction: either all of the previewed records are added or none are. Existing profile details are never overwritten, and imported experience and education stay unverified until you review them in your profile.
- Review the imported profile facts and mark the accurate ones as verified.
- If anything is wrong, restore the backup from the start of this procedure
(
npm run restore -- <backup-file> --yeswhile ReboundOps is stopped). - Delete the staging folder when you no longer need it. It contains personal data.