2026-09-01 - obSCEne oracle mined out for clean fixes; pivot to a real title next
Characterised the four remaining obSCEne failures under orbistoun and
found none is a clean, safe fix (D448): 900-surface/control
is the deliberate loud-stub-everything model (D392) - a correct narrow
fix exists (refuse weak+unknown imports, matching hardware's
weak-undefined-null) but needs an ELF binding decode + service
refuse-set and re-baselines every measurement, so it is a deliberate
deferred step, not a loop flip; 110-modules/info-size and
/names need the SceKernelModuleInfo layout
obSCEne itself says it does not fully know (uncitable, D008);
137-kernelcall/system-version skips on hardware (getpid
gadget) and is the same resolve-everything model plus honest ENOSYS on a
raw syscall. Five real fixes came out of this oracle (D443-D447 +
flexible); the rest are design or layout-gated. Next frontier is a real
title: PPSA02664 runs 1541 calls and walls in
libc::_Getpctype (the ctype-table accessor), whose table
FreeBSD documents.