Regenerate the published doc pages #218

2026-09-22T17:46:39Z

paulhitt commented

2026-09-22 17:46:39 +00:00

(Migrated from gitlab.com)

doc-pages-head has been red on main. flamenet-e2e moved and the pages this relay serves were never rebuilt, so the site has been describing a repository that changed underneath it.

What the public pages were saying:

page	said	is
`spec.leaf`	50 client tests, 16 relay, 24 interop	167, 265, 92
`spec.leaf`	`test/client.mjs` as "27 of the 77" and "51 of the 113"	one file, one total
`spec.leaf`	a §5.4 cross-reference	§5
`audit.leaf`	five suites under `engine/`	seven under `test/`, now named
`engine.leaf`	the README as it was	current

The counts understated the test surface rather than overstating it, which is the better direction to be wrong in. But audit.leaf exists to tell an auditor where to look, and it pointed at a directory that is not there.

Note for review: the §19.5 and §20.6 text is unchanged by this. Those sections are byte-identical before and after; the published page already matched its source there. The staleness was counts, cross-references and the test layout.

.e2e-ref goes to 32f01d1. Part three of the procedure in .e2e-ref.md, "make whatever source change the new conformance suite expects", is empty this time: conformance-head is already green at that commit.

The pages were generated from a checkout pinned at exactly 32f01d1 rather than from a working copy, so they cannot have been built from anything but what this commit pins. check-doc-pages: 9 generated pages match their sources. Secret sweep clean.

Generated with Claude Code.

`doc-pages-head` has been red on main. flamenet-e2e moved and the pages this relay serves were never rebuilt, so the site has been describing a repository that changed underneath it. What the public pages were saying: | page | said | is | |---|---|---| | `spec.leaf` | 50 client tests, 16 relay, 24 interop | 167, 265, 92 | | `spec.leaf` | `test/client.mjs` as "27 of the 77" and "51 of the 113" | one file, one total | | `spec.leaf` | a §5.4 cross-reference | §5 | | `audit.leaf` | five suites under `engine/` | seven under `test/`, now named | | `engine.leaf` | the README as it was | current | The counts understated the test surface rather than overstating it, which is the better direction to be wrong in. But `audit.leaf` exists to tell an auditor where to look, and it pointed at a directory that is not there. Note for review: the §19.5 and §20.6 text is **unchanged** by this. Those sections are byte-identical before and after; the published page already matched its source there. The staleness was counts, cross-references and the test layout. `.e2e-ref` goes to `32f01d1`. Part three of the procedure in `.e2e-ref.md`, "make whatever source change the new conformance suite expects", is empty this time: `conformance-head` is already green at that commit. The pages were generated from a checkout pinned at exactly `32f01d1` rather than from a working copy, so they cannot have been built from anything but what this commit pins. `check-doc-pages`: 9 generated pages match their sources. Secret sweep clean. Generated with Claude Code.

paulhitt commented

2026-09-22 18:28:25 +00:00

(Migrated from gitlab.com)

mentioned in commit 6ca295a654

mentioned in commit 6ca295a6546e65371666d097c8cd6ce7c83b26cc

paulhitt (Migrated from gitlab.com) merged commit 6ca295a654 into main

2026-09-22 18:28:25 +00:00

Sign in to join this conversation.

1 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: extant2000/flamenet-relay#218