AnilAli0201 · 30 Nov 2025, 14:00 UTC
Keep the identified parts and full source images, with their capture conditions. Exclude unsupported scratch labels from tuning, but retain their history so the next person can see why the working set changed.
15 pointsOriginals retained. The working folder now separates confirmed surface marks from unverified image labels.
0 pointsDoes the faint rejected bezel appear in the new comparison too, or only the easier-to-see marks?
13 pointsIt is included. Not the same part from the FR10 work; quality identified this one separately.
8 pointsThen compare its repeat presentations alongside an accepted difficult surface, keeping the part identities attached to every capture. A set of obvious scratches alone will give you a very comfortable but incomplete answer.
9 pointsCan the next loader tell when the bezel is seated correctly? I do not know your holder, but a small unnoticed tilt could keep making new-looking examples.
9 pointsOmar, worth checking before they tidy the folder again; the loader needs something more concrete than make it look like this photograph.
24 pointsOne locating edge is hard to see. Tooling is reviewing that with the loader; tuning remains paused.
6 pointsThis discussion is closed to new replies after six months without activity. Last activity: .