backlog: rule G26 sittings 1 and 2 (33 rows), file #1211 - #312
Merged
Conversation
Sittings 1 and 2 of the standing demand-gate triage, ruled under the owner's delegation. 33 rows resolved: 18 DROP, 12 deferral confirmed with a dated review, 3 promoted out of demand-gate. Open items 209 -> 191, then 192 with #1211 filed. WHY DROPPING IS THE HONEST OUTCOME, not neglect. Every drop was verified to have a working alternative in shipped code, or a trigger that cannot arise as the engine is built. Re-filing a dropped row costs one item; carrying it costs a re-read and a re-price at every planning pass. Recording a row as "not yet triggered" year after year, with no evidence the question was ever re-asked, IS the failure mode. The line applied, so it can be re-applied to the remaining sittings: KEEP the rows whose trigger, if it fires, BLOCKS a feed; DROP the rows whose trigger merely INCONVENIENCES, because a shipped mechanism, a configuration change or the network layer already answers it. THREE DROPS ARE STALE RATHER THAN PENDING, and they are the ones worth noticing -- in a status scan a demand-gate row whose trigger has ALREADY BEEN SERVED looks identical to one still waiting: #127 proxy auth -- fired and answered; #112/#128 shipped under ADR 0126 and the NTLM residual is refused at construction with the cntlm workaround named. Duplicates an ADR decision. #125 uploaded logs -- fired and served across five API routes and seven /ui routes, none of which ingests into the live store. #130 shared queues -- cannot fire: it describes a competitor's internal queue model, not a capability a feed here lacks. PROMOTED, each because its API half already ships and only the console half is absent: #177 (no PHI), #132 (trigger fired against this repo's own sample), and #124 -- a PHI EGRESS PATH, which must ship behind the same step-up and scope controls as its API half or it is a downgrade. FOUR ROWS ROUTED TO THE OWNER rather than ruled: #141 and #178 are policy questions about the cipher floor and out-of-engine relays; #158 is a promotion argued on security-sequencing rather than demand; #105 is gated on #313, which does not exist in the public namespace and is resolvable only internally. ALSO FILES #1211, allocated with scripts/coord/alloc.ps1 -- empty_claims_per_msg's contention-immunity claim measured false on a hosted runner. The confirming re-run was predicted BEFORE it ran, so its green is a result and not a rationalisation. Its mechanism section records what the excursion is NOT, because the near miss was recalibrating an SLO on a causal story the same test run already refuted. THE LEAK GATE BLOCKED THE FIRST ATTEMPT AT THIS COMMIT AND IT WAS RIGHT. #169's rationale named the vendor adjacent to the estate: the vendor name alone is ordinary competitor prose and appears throughout this file, but paired with the estate word it implies a specific customer's installation. Isolated by probing the scanner phrase by phrase rather than by reading the token out of a log, and fixed by removing the adjacency -- not by an allowlist entry, which the gate's own remediation text explicitly refuses for a real token. Ledger, re-derived with parse_items: live 242, open 192, closed-in-live 50, archive 236, namespace 478 conserved. All three ledger gates pass.
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.
Rules G26 sittings 1 and 2 over the DEMAND-GATE surface. Ledger only -- no engine, IDE or webconsole
code is touched.
BACKLOG #1211
What it does
33 rows resolved: 18 dropped, 12 deferrals confirmed with a dated review, 3 promoted out of
demand-gate. Roughly two thirds of the demand-gate surface, with no engineering.
BACKLOG #1211is filed by this commit:empty_claims_monotonicexcursions exceed the band its ratioform claims to be immune to. Its scope says to establish the true variance from several samples at one
N, then either widen the band to the measured distribution -- recording the pool and date as data -- or
replace the monotonicity expectation with one the metric can support. Explicitly not a flaky marker,
which would re-hide the class. It also records what the excursion is NOT, preserving the
fd_count_monotonicdiscriminator so the next reader does not re-derive a causal story that wasalready refuted.
Verification
parse_itemsfrom this branch rather than carried forward: live 242, open192, closed-in-live 50, archive 236 -- namespace 478 conserved. No duplicate numbers, and no item
declaring more than one status.
main; the instrument was checked first --merge-tree --write-treeagainst aknown-conflicting branch returns exit 1, so exit 0 here means clean rather than blind.
#1040,#1210and#64confirmed still open; none was tidied.than a vacuous green on an empty ruleset.
Note
Four rows are routed to the owner for a ruling rather than decided here -- two policy questions about
the cipher-floor posture, one promotion argued on path-traversal grounds rather than demand, and one
whose stated gate is an internal item number that does not exist in the public namespace.