)]}'
{"/COMMIT_MSG":[{"author":{"_account_id":1179,"name":"Clay Gerrard","email":"clay.gerrard@gmail.com","username":"clay-gerrard"},"change_message_id":"cbbdcd91c81ab13e76f07f8d85b120a0ce3f693e","unresolved":true,"context_lines":[{"line_number":10,"context_line":"no user metadata stored in the clear. Metadata written before encryption"},{"line_number":11,"context_line":"was deployed, or while disable_encryption is set, stays in the"},{"line_number":12,"context_line":"x-object-meta- namespace instead of the transient sysmeta crypto"},{"line_number":13,"context_line":"namespace, and such an object must not be reported as encrypted."},{"line_number":14,"context_line":""},{"line_number":15,"context_line":"Co-Authored-By: Claude Opus 5 \u003cnoreply@anthropic.com\u003e"},{"line_number":16,"context_line":"Change-Id: I2fe8eb15578c58faa00a3e6fb424ef990466a432"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":1,"id":"c3f2272b_32884585","line":13,"updated":"2026-08-21 05:40:49.000000000","message":"FWIW a non-zero object with encrypted data that later recieves a POST while encryption is disabled will still have valid/complete put_crypto_meta and report the cipher to indicate \"the body is encrypted\"\n\nIMHO the problem is not that un-encrypted metadata should be driving the returned backend cipher header\n\nThe \"problem\" is that content-length\u003d0 is a *terrible* proxy for the \"encrypted policy\" when the data was written:\n\n1001751: encryption: persist body cipher policy | https://review.opendev.org/c/openstack/swift/+/1001751","commit_id":"90b376dd8ad45057fc78ec7d4e1d96ef0d5b0879"}],"/PATCHSET_LEVEL":[{"author":{"_account_id":1179,"name":"Clay Gerrard","email":"clay.gerrard@gmail.com","username":"clay-gerrard"},"change_message_id":"cbbdcd91c81ab13e76f07f8d85b120a0ce3f693e","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"e210dfbe_8486613f","updated":"2026-08-21 05:40:49.000000000","message":"this is probably better than not and may be sufficient - but I think we should stop trying infer encryption policy of the body at the time of write from huristics and should instead just explicitly write it down and only claim what we can PROVE.","commit_id":"90b376dd8ad45057fc78ec7d4e1d96ef0d5b0879"}]}