Backlog

Known gaps, loosely ranked. Everything here is a thing this program should be able to say and currently cannot.

Written down rather than left implicit, because "we didn't check that" and "we checked and it was fine" look identical in a report that simply has no entry. This table is generated. Edit an item under backlog/, then run tools/split-doc.sh --index obscene BACKLOG 2 backlog.

	item	status
⚪	Where it stands	no marker
⚪	11. Relational checks, beyond the seventeen that exist	no marker
⚪	12. N-way consensus as a substitute oracle	no marker
🟢	0. The vendor dynamic table	done
🟢	0b. Reading a report out of an emulator	done
⚪	0c. The second and third emulators - both now run, one still resolves nothing	no marker
⚪	1. The current generation's graphics interface - censused, and now called	no marker
⚪	2. Struct-taking functions are absent from every behavioural check	no marker
🟢	3. Verifying the NID mapping	done
⚪	4. Libraries never listed at all	no marker
⚪	4b. setjmp / longjmp	no marker
⚪	5. The positive-to-negative ratio is still wrong	no marker
⚪	6. Build speed over the mount	no marker
⚪	6b. pretty does not render the responsiveness records	no marker
⚪	6c. The thread-churn crash is measured but not characterised	no marker
⚪	6d. Condition variables and barriers - the non-blocking half is checked	no marker
⚪	7. The census cannot detect a lying platform - now it says so, in the report	no marker
⚪	8. Coverage: 312 names are still presence-only	no marker
⚪	9. Not under version control	no marker
⚪	10. The gap the emulators measured	no marker
⚪	13. Getting the report off the machine	no marker
⚪	14. Measurements orbistoun is blocked on	no marker
⚪	15. Two probes for the heaviest blocked premises: POSIX error encoding, and sync bounds	no marker
⚪	16. The futex timeout, and what the two threading probes still owe	no marker

	meaning
🟢	done
🟡	begun
🔴	open, or explicitly not done
⚪	deferred, not planned, or carrying no marker either way