probe: does a four-segment URL publish on this site? - #83
Merged
Conversation
Spec 010's S3 caps a published path at three segments on the stated grounds that three is the deepest the live site is KNOWN to work at and four was never tested. That is an evidence boundary, not a platform limit, and it is now costing navigation quality in two places: MigratingToPollyV8.md cannot nest under its own source, and AzureBlobConfiguration.md cannot sit under AzureBlobArchiveProvider.md under a new OutboxArchiver.md. GitBook's own documentation publishes 30 pages at four segments below its site root (e.g. create-content/content-structure/page/tags), so the platform supports the shape. This measures our site. A brand-new page at a path that has never existed is used deliberately, so no automatic redirect can mask the result -- D0's reasoning. No existing page moves, so no real URL churns and no redirect caches against content a reader wants. Predicted: guaranteed-at-least-once/azureblobarchiveprovider/azureblobconfiguration/urldepthprobe To be reverted immediately after the measurement is recorded. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_012tcdwxVb8NmKaX2S6fvyFg
iancooper
added a commit
that referenced
this pull request
Aug 8, 2026
Four segments publishes correctly on this site. Measured against the live site 2026-08-08, minutes after PR #83 merged: path guaranteed-at-least-once/azureblobarchiveprovider/ azureblobconfiguration/urldepthprobe sitemap present; 112 entries, was 111 status 200 location 0 headers -- no redirect, this is a genuine page body 529,080 bytes (a 404 shell is ~189.5 KB) title URL Depth Probe | Paramore Brighter Documentation S3's ceiling of three was an untested assumption, not a platform limit. GitBook's own documentation already published 30 pages at four segments below its site root; this establishes the same for our site. Reverts the content of 39823ea (PR #83). The tree is byte-identical to c4aedb5, the commit before the probe. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_012tcdwxVb8NmKaX2S6fvyFg
iancooper
added a commit
that referenced
this pull request
Aug 8, 2026
…as distorting
S3 capped a published path at three segments on the stated grounds that
three is the deepest the live site is KNOWN to work at. Read closely that
is a fact about our own SUMMARY.md -- nine pages happen to be nested one
level, none two -- not about GitBook. It had already bent two placements
out of shape:
- MigratingToPollyV8.md had to become a SIBLING of its own source
- AzureBlobConfiguration.md could not sit under AzureBlobArchiveProvider.md
under the new OutboxArchiver.md
Measured instead of assumed, 2026-08-08:
- GitBook's own documentation publishes 30 of 182 pages at four segments
below its site root (create-content/content-structure/page/tags), which
is section + two ancestor pages + page -- exactly our shape
- PR #83 published a NEW page nested three levels deep; a path that had
never existed, so no automatic redirect could mask it. Present in the
sitemap (112 entries, was 111), 200, ZERO location headers, 529,080-byte
body against a ~189.5 KB 404 shell. PR #84 reverted it; sitemap back to
111 and the tree byte-identical to c4aedb5
- urlmap.py needed no change: its `[<ancestor>/]*` was never bounded
design.md: S3 amended to <=4 in §4, §7.6's superseded reasoning quoted and
retired in place, and §17 records the measurement and what moves.
tasks.md: MigratingToPollyV8.md nests under its own source at four
segments; OutboxArchiver.md and TransactionalMessagingWithTheOutbox.md go
top-level; AzureBlobArchiveProvider.md and its configuration child
re-parent under the archiver. Outbox and Inbox lands at 9 top-level
entries. Task 2.2's --check-shape asserts 4, and its prove-it-fails step
now nests to five.
The cost, stated not buried: those two Azure Blob pages are the only two
whose URL moves twice, so Task 5.1 owes two redirect entries for the
intermediate paths. Design §7.6's no-page-moves-twice property holds for
the other 140.
S3 is the FIRST RULE in this programme to move; all fourteen previous
corrections were tallies. It was not wrong so much as over-restrictive,
and it survived requirements, design and a design review unchallenged.
When a rule's justification is the absence of evidence, that is a
measurement waiting to be taken.
Appendix A re-verified: 32 rows, 5 top-level, 26 at depth 3, 1 at depth 4;
section allocation still matches design §7.6 row for row; 142 pages.
No file under contents/ changed -- linkcheck 112, pagelint 0 errors.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_012tcdwxVb8NmKaX2S6fvyFg
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.
Temporary measurement for Spec 010. To be reverted immediately after the result is recorded.
Spec 010's shape rule S3 caps a published path at three segments, on the stated grounds that three is the deepest the live site is known to work at and four is unverified. That is an evidence boundary, not a platform limit — and it is now costing navigation quality in two places:
MigratingToPollyV8.mdcannot nest under its own source, becausePolicyRetryAndCircuitBreaker.mdalready publishes at three.AzureBlobConfiguration.mdcannot sit underAzureBlobArchiveProvider.mdunder a newOutboxArchiver.md.GitBook's own documentation publishes 30 pages at four segments below its site root (e.g.
create-content/content-structure/page/tags— section plus two ancestor pages plus page), so the platform supports the shape. This measures our site.Method
A brand-new page at a path that has never existed, so no automatic redirect can mask the result — D0's reasoning. No existing page moves, so no real URL churns and no redirect caches against content a reader wants.
Predicted path:
Read from
sitemap-pages.xmland the response fingerprint, never the status code — every cached response on this site reports200.🤖 Generated with Claude Code
https://claude.ai/code/session_012tcdwxVb8NmKaX2S6fvyFg