Status: design proposal, no implementation, no measurement. Origin: Travis, 2026-09-06, in a Codex session — "what do you think about designing a sort of FPGA but instead a sort of Floor Programmable Locus Array?" The research-and-design notes that session produced are kept here verbatim as DESIGN_NOTES_2026-09-06.md (source: ~/Documents/Codex/2026-09-06/wha/outputs/), so the concept has a revision in this tree instead of living only in an app's output folder.
Where it sits. FPLA is a contract the existing software floors can implement first (ComputeDriven's Locus, floor, placement, readiness and currentness semantics — computedriven/); hardware is a later implementation. The notes' first experiment — a reference machine for two independent service elements, compared transition for transition with the software reference — is the same artifact as ComputeDriven's next ruled step, the compiled home (computedriven/receipts/ R7.2.1A3-OPEN.md §5.6). Progress on FPLA is recorded in ../FOUNDATION_LANE.md.
Naming. FPLA already means Field-Programmable Logic Array in the hardware literature; use "ComputeDriven FPLA" or the expanded name outside this tree.
Not defined here: the "Transistor" concept. It has no definition on this machine as of 2026-09-08 and is requested from Travis.