# The two decoder eras, rendered: all 16 composable bundles of GP_TITLE. # 2026-08-30. # # STALE era = origin/main (ui_layout.rs: "Keyframe time, or None for the # group's last frame" -- a pose's time read from the NEXT record) # FIXED era = formats-pin-2026-08-30 == this branch (a pose's time PRECEDES it) # # Both built from source; renders are `screen render --all --build N --primitives`. # # CONTROL 1 -- the two binaries really do embody the two eras: # stale rest (0,0) t=70 [12:0,0 70:0,0 80:0,0 -:0,0] # fixed rest (0,0) t=12 [0:0,0 12:0,0 70:0,0 80:0,0] (build 5, pteff00.prm) # CONTROL 2 -- the renderer is deterministic: same binary, same flags, twice, # entry 7 -> 0 differing px, entry 12 -> 0 differing px. entry what differing px RMSE max|d| 0 loading 0 0.000 0 1 loading 0 0.000 0 2 plate 0 0.000 0 3 plate 0 0.000 0 4 title 0 0.000 0 5 main_menu 0 0.000 0 6 extras 0 0.000 0 7 title_jp 74507 12.409 233 <-- DIFFERS 8 main_menu_jp 0 0.000 0 9 extras_jp 0 0.000 0 10 splash publisher 32842 1.767 35 <-- DIFFERS 11 splash developer 23201 0.942 31 <-- DIFFERS 12 loading (dressed) 49771 10.078 229 <-- DIFFERS 13 splash publisher twin 31708 1.782 35 <-- DIFFERS 14 splash developer twin 23201 0.942 31 <-- DIFFERS 15 loading twin 49771 10.078 229 <-- DIFFERS 7 of 16 bundles differ between the eras; 16-7 are byte-identical. # MECHANISM, entry 7 (title_jp): exactly ONE element's rest POSITION moves. # stale 8 ptlogo_eff3.t32 ... rest (108,72) # fixed 8 ptlogo_eff3.t32 ... rest (98,42) # ptlogo_eff3.t32 is the element MISSION.md and ui-resting-pose.md already # name as THE plateau-less rest() discriminator. 74 507 px, RMSE 12.4. # MECHANISM, entries 10-15: NO rest position changes at all. The rest # SELECTION moves to a different keyframe which happens to sit at the same # (x,y) with a different SCALE and ALPHA -- e.g. pgloading_delta.t32 holds # (120,560) at 0%,0% a=0 / 75%,75% a=128 / 96%,96% a=192. Comparing the # 'rest (x,y)' column alone says nothing changed; the pixels say otherwise.