Features/Managed sync

• Managed storage & sync
  • Known to be buggy. Tine-managed storage does not yet fully work in our own testing; we're actively working on it. Use it only on a graph you are comfortable testing. Your normal mode is Direct files: Tine reads and writes your existing Markdown or Org graph, and Syncthing, Dropbox, Logseq, and other tools can keep synchronizing those files exactly as before. Nothing changes unless you explicitly enable this mode for one graph.
• What the opt-in does
  • Tine stores operation history in that graph's .tine-sync/ directory while keeping your Markdown/Org files in place as an editable projection. Logseq and other file-based tools can still read those files; Tine conservatively reconciles outside edits.
  • This is a separate mode, not a replacement for ordinary Syncthing/Dropbox. Do not enable it merely because you already synchronize your graph folders.
• Current scope
  • Managed storage currently covers page and journal text only. Assets, PDF sidecars, and configuration remain ordinary provider-synchronized files.
  • Setup first makes a safety backup and checks that Tine can reconstruct the same Markdown/Org tree. If that cannot be proven, it stops instead of changing graph authority.
• Review a detected group deletion
  • This safety surface belongs only to a graph using Tine-managed storage. Direct-files graphs keep their ordinary file and snapshot recovery paths.
  • Four deleted pages together become a Tier 2 sweep. A deletion reaching at least 10% of the graph's pages (rounded up), or 50 pages when that is lower, becomes Tier 3. Tier 1 stays quiet. A Tier 3 sweep also pauses propagation for about five minutes so another device is less likely to carry a mass deletion onward before you can review it.
  • Tine shows a warning and keeps a Deleted pages dock with the sweep's member pages and live action status. Closing either the warning or the panel makes no decision and deletes no sweep record.
  • Choose Restore to revive every member from its accepted predecessor state, Re-apply to accept and re-author the deletion against the current state, or Keep deletion to deliberately retain the accepted deletion. These are the only actions that dispose the sweep.
  • Restore may run in several durable chunks. The panel shows its progress; if it fails, it keeps the recorded cause and offers Run Restore again to resume through the same backend recovery path.
  • What you should see: the dock remains until you make an explicit choice, and then keeps the completed disposition visible as history. See Reference/Troubleshooting and recovery for the same steps in recovery form.
• Try it on a test graph
  • 1. Open Settings (t s) → Backups & recovery → Storage & sync.
  • 2. Read the Known to be buggy warning, then choose Enable Tine-managed storage... for a graph you are comfortable testing.
  • 3. Wait for setup to finish. Your Markdown/Org files remain beside the new .tine-sync/ data; do not edit the internal sync directory yourself.
  • 4. To prepare a second test device, first let your existing provider deliver the same test graph folder there. Then use Set up sync with another device... on the first device and Join this synced graph... on the other one. Tine verifies that both devices are joining the same graph history.
  • 5. What you should see: Direct files remains available for ordinary graphs, and managed storage is enabled only for the graph where you chose it.
• Return to Direct files
  • The same panel offers Return to Direct files when Tine can first preserve complete recovery state. If it says that safety could not be verified, leave the graph alone and resolve that state rather than forcing a file-mode switch.