Summary
Revision Control (Lore) fails on one UEFN project with a misleading “NotFound” error. Lore.log shows the real cause one step earlier: a fragment fails Zstd decompression/verification — i.e. a corrupted staged-change fragment stored server-side, left over from a crash during a revision-control operation. A second repository on the same account is unaffected and syncs normally, which confirms the corruption is isolated to one repository’s staged state, not the local client. This appears to be the same issue tracked as FORT-1129923.
Please select what you are reporting on:
Unreal Editor for Fortnite
What Type of Bug are you experiencing?
Lore
Steps to Reproduce
- Have a UEFN project connected to Revision Control (Lore).
- A crash occurs during a revision-control operation (e.g. check-in), leaving a corrupted staged change.
- Relaunch UEFN and query status / attempt to reconnect on that project.
Expected Result
Revision Control reconnects and reports status successfully, or recovers gracefully with a clear, actionable error (e.g. an option to discard the corrupt staged change).
Observed Result
Any status query that reads staged state on the affected repository fails. UEFN surfaces:
Failed to deserialize revision states: Not found | NotFound(Traced { inner: NotFound, trace: Trace { locations: [Location { file: “lore-revision\src\state.rs”, line: 546, column: 37 }] } })
Lore.log shows the real cause one step earlier:
[lore_storage::compress] Zstd decompress failed: 18446744073709551606
[lore_storage::read] Fragment 8f291f20faed0af2656d76f54e29b3c65fda80303dea3bf7c5392ab31a34245c failed decompression/verification: failed to decompress fragment
It occurs specifically when querying staged status (staged:1, sync_point:1) on the affected repository. A separate repository under the same identity completes status successfully (branch synchronized, revision 81), confirming the failure is server-side and isolated to the corrupted staged fragment. The affected repository’s last committed revision is intact — only the uncommitted staged state is corrupt.
Platform(s)
PC / Windows
Additional Notes
Details for the team
Identity: 01975155bdd171c995aedfaecd20aa30
Affected repository: 450029d04993f1f1e0cfdb81cdf0c8a7
Healthy repository (reference): 809c96e55e5b37d6810fcda2fbce8de6
Corrupted fragment: 8f291f20faed0af2656d76f54e29b3c65fda80303dea3bf7c5392ab31a34245c
Lore version: 0.8.1-nightly+3397
Backend: urc-uefn.live.ucs.on.epicgames.com
Related report: FORT-1129923
Note on local resets
Since the fragment is stored server-side, local mitigations (clearing caches, reinstalling UEFN, deleting the local .lore folder) are not expected to resolve this — consistent with the existing report FORT-1129923, where a full local reset had no effect.
Request
As the corrupted fragment is an uncommitted staged change and the last committed revision is intact, please discard/purge the corrupt staged change on repository 450029d04993f1f1e0cfdb81cdf0c8a7 (fragment 8f291f20…a34245c) so the client can resync from the last good revision. Full Lore.log attached.
Uploading: Lore.log…