Execute FABRIC-3.md §XXXV.6: format new artemis.img, re-mint all 9 identities live
Build / build-amd64-iso (push) Canceled after 0s
Build / build-aarch64-iso (push) Canceled after 0s
Build / build-riscv64-img (push) Canceled after 0s

Boots amd64 with Zuse's bleached thumbdrive, confirms live: fresh 1 GiB
disk/artemis.img formats ("Artemis: blank disk -- formatting"), Zuse
genesis-mints onto it ("Zuse: genesis minted onto attached thumbdrive",
root CA written to the new fence's devblock 0), session authenticates
(zuse@Hera ok>). With Zuse live, mints the other 8 identities (bob, 00-06)
one at a time via QMP-driven xHCI hotplug + MINT at the console, each
confirmed by its own "MINT: identity minted" log line before detaching
and moving to the next. Clean BYE shutdown, zero leaked qemu/socat
processes, all 9 images verified nonzero.

Root cause confirmed live before executing: Zuse's genesis marker lives
on artemis.img's own fence (devblock_from_top=0), not on her thumbdrive
-- with the marker gone, her old drive could not have re-authenticated
regardless of its own content, confirming re-mint was genuinely required.

FABRIC-3.md §XXXV.7/.8, disk/README.md updated with the full account.
Serial log preserved: logs/20260917-114652/amd64/.

Still open: the block-backed DoE persistence writer itself is unbuilt
(design only); ZUSE-ELIGIBILITY-ADD (elevation eligibility, narrower
than baseline auth) not re-populated for any of the 8; old 30 MiB
artemis.img backup disposal undecided.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
This commit is contained in:
Robert Allan James
2026-09-17 11:57:10 -04:00
co-authored by Claude Sonnet 5
parent e56974e0bf
commit 3223fac2a0
13 changed files with 9354 additions and 9 deletions
+28 -3
View File
@@ -18,9 +18,21 @@ VM's `--disk-img=` flag instead.
ratified building a brand-new blank image and re-minting Zuse + all
thumbdrive identities fresh against it, since re-minting invalidates the
prior identities regardless of whether a migration is attempted. Blank
(all zero) at creation, same as every other fresh-format fixture below —
not yet formatted or re-minted; that is a separate, later live-boot step.
The superseded 30 MiB image is preserved, not deleted, as
(all zero) at creation, same as every other fresh-format fixture below.
**Formatted and re-minted live, same day**: booted amd64 with Zuse's
(freshly bleached) thumbdrive attached at launch — `Artemis: blank disk --
formatting` confirmed the fresh format, `Zuse: genesis minted onto
attached thumbdrive` confirmed her genesis mint (root CA — her
`zuse_pubkey` — written to this image's own devblock 0, per
`capsule_zuse_boot.c`'s `genesis_marker_read()`/`install_and_activate()`),
and the `zuse@Hera ok>` prompt confirmed her session authenticated. The
other 8 identities (`bob`, `00`–`06`) were then minted one at a time via
QMP-driven xHCI hotplug (`drive_add`/`device_add` per drive, `MINT` or a
`MINT-NUM` helper word at the live console, `device_del`/`drive_del` to
detach before the next) — all 9 `MINT: identity minted` confirmations
present in `logs/20260917-114652/amd64/qemu-amd64-20260917-114652.log`,
clean `BYE` shutdown, no leaked qemu/socat processes. The superseded
30 MiB image is preserved, not deleted, as
`artemis-30mb-pre-1gib-backup.img` (below).
- `artemis-30mb-pre-1gib-backup.img` — the pre-2026-09-17 `artemis.img`
(30 MiB), preserved rather than deleted when `artemis.img` was rebuilt
@@ -123,6 +135,19 @@ carrying timestamp noise in git history.
default via zeroed former padding, not corruption) rather than crashing
or misreading adjacent fields.
- `thumbdrives/` — the 9 **real** identity thumbdrive images (`zuse-thumb-
ident.img`, `bob-thumb-ident.img`, `00-thumb-ident.img`–`06-thumb-
ident.img`), 64 MiB each, minted via the actual kernel `MINT` word (not
blank placeholders). **Bleached and re-minted fresh, 2026-09-17**
(FABRIC-3.md §XXXV.6/§XXXV.8): required alongside the `artemis.img`
rebuild above, since Zuse's own genesis marker lives on `artemis.img`,
not on any thumbdrive — with that marker gone, none of these 9 images'
prior certs could authenticate against the new disk regardless of their
own on-drive content, and `capsule_mint_identity()` refuses to mint over
an already-recognized drive, so all 9 had to be blanked first. Same
verification convention as the original 2026-09-06 mint
(`project_thumbdrive_identities_minted_20260906.md`): nonzero-byte count
in each image's first 6 devblocks confirms real identity data present.
- `zuse.img` — 64MB raw image simulating the physical Zuse superuser
thumbdrive for QEMU testing (FABRIC-2.md, Phase 8: `zuse.img` "bleach"
mechanism, added 2026-08-26). **64MB is only this fixture's size, not a