jan/rrt
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Classify machine shop Tier-2 selector outlier

Browse source
This commit is contained in:
Jan Petykiewicz 2026-04-19 15:54:32 -07:00
commit 7894c52aea
5 changed files with 6714 additions and 278 deletions
Download patch file Download diff file Expand all files Collapse all files

8
docs/control-loop-atlas/map-and-scenario-content-load.md
View file

@ -108,6 +108,14 @@
`MachineShop` inside the `TextileMill` cluster (`byte_0xba = 0x3f`, `byte_0xbb = 0x00`). So the
next load-side source-selection pass should focus on that row-level outlier and any matching
replay/seed logic, not on a whole-cluster nonzero bank hypothesis.
The direct-name plus alias plus selector join sharpens that one more step: inside the same
nonzero `0x000001f4` family, the direct `Warehouse/TextileMill/Warehouse` shape splits into six
all-zero selector peers (`ConcretePlant`, `ConstructionFirm`, `ElectronicsPlant`, `Hospital`,
`PharmaceuticalPlant`, and `Warehouse`) plus one unique selector outlier,
`MachineShop = 0x00/0x80/0x3f/0x00`. The sibling bare `Port/TextileMill/Port` row stays
all-zero. So the remaining load-side question is no longer whether the bare `Port` / `Warehouse`
row carries the stock exception; it is why one warehouse-shaped industrial peer in that same
alias family carries the lone selector while the bare rows do not.
The global stock selector report tightens that further: the full `MachineShop.bca` signature
(`0x00/0x80/0x3f/0x00` across `0xb8..0xbb`) is unique across the checked-in stock `.bca`
corpus. So the remaining load-side Tier-2 frontier is one surfaced stock-file outlier plus the