NA-0684 — the reference runbook stops naming retired operator hostnames - #42
Merged
Conversation
…r hostnames Five occurrences in docs/NA-0004_reference_deployment_runbook.md become angle-bracket placeholders: <attachments-public-host> on the reference-profile line, the TLS-termination line, the Caddyfile block and the public probe; <relay-public-host> on the mandatory relay preflight. Command structure is unchanged, one substitution per line. AUTHORITY: the operator disposition of 2026-07-28 recorded in spine D-1320's follow-up map -- the retired public hostnames are operator-owned, retired from service, and their registrations are HELD INDEFINITELY. Live instructions must stop naming them; dated history stays intact. No history is rewritten anywhere: the mitigation for what git history already carries is the registration hold, a standing operator action, not scrubbing. THE PROPERTY THAT DECIDES WHAT MOVES, ruled by the Director and recorded verbatim in the spine decision: a line is in scope when, read today, it DIRECTS TRAFFIC -- a default, a command, a configuration a reader copies; a line is out of scope when it REPORTS WHAT WAS TRUE -- a measurement, a decision, an evidence capture. So the runbook is fixed and the NA-0003 / NA-0004 / NA-0005 evidence files are NOT: eight occurrences in tests/ are left byte-identical, and the lane's gate prints them in every run so the exception cannot hide inside a green result. THE SSH ALIASES STAY, and that is a measured finding rather than a preference: qatt is a PROJECT TERM -- it names the canonical attachment-service contract DOC-CAN-006 and appears in this crate's own source -- and qsl is the project. Only the DNS names are operator infrastructure. Census for this repo: 13 occurrences over 41 tracked files / 13764 lines, split 5 live instruction + 8 dated record, with the classes asserted to sum to the raw count. infra-literal-scan (staged): clean, 2 files, 33 lines. (tree): clean, 41 files, 13792 lines. Note the gate CANNOT see this class -- the retired names are in no digest list and match no structural pattern -- which is why the lane brought its own instrument.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Goals: G4, G1
Five occurrences in
docs/NA-0004_reference_deployment_runbook.mdbecome angle-bracketplaceholders —
<attachments-public-host>on the reference-profile line, the TLS-terminationline, the Caddyfile block and the public probe;
<relay-public-host>on the mandatory relaypreflight. Command structure is unchanged; one substitution per line.
Spine directive D619 (
a8dab7f1…7ea65092, 539 lines), lane NA-0684, repo-localD-0013.
The authority, and the property that bounds it
The retired public hostnames are operator-owned, retired from service, and their
registrations are held indefinitely (operator, 2026-07-28, in D-1320's follow-up map). Live
instructions must stop naming them; dated history stays intact, and no history is
rewritten — the mitigation for what git already carries is the registration hold, not
scrubbing.
That is why the runbook changes and the
NA-0003/NA-0004/NA-0005evidence files donot: 8 occurrences in
tests/are left byte-identical, and the lane's gate prints themin every run so a leave cannot hide inside a green result.
⚠ The ssh aliases stay — measured, not preferred
qattis a project term: it names the canonical attachment-service contract(
DOC-CAN-006) and appears in this crate's own source;qslis the project. Only the DNSnames are operator infrastructure. This closes "should the aliases go too?" with evidence.
Impact
Documentation only. No source, no test, no workflow, no dependency, no behaviour. The runbook
still reads as a runbook — a reader supplies their own host where the placeholder sits.
No-regression
2 files: the runbook (5 substitutions) and
DECISIONS.md(D-0013). Nothing else in thisrepository is touched. Census for this repo: 13 occurrences over 41 tracked files / 13 764
lines, split 5 live-instruction + 8 dated record, with the classes asserted to sum to
the raw count — a lost hit crashes the instrument rather than being miscounted.
Tests-Vectors
infra-literal-scan --mode staged: clean (2 files, 33 lines).--mode tree: clean (41files, 13 792 lines). ⚠ Note the committed gate cannot see this class — the retired names
are in no digest list and match no structural pattern — which is exactly why the lane brought
its own instrument, run RED first at base. No code paths touched, so there are no vectors
to add.