Say how CI reads the stale-preview class, since it cannot capture the stills - #175
Say how CI reads the stale-preview class, since it cannot capture the stills#175leaf-agent wants to merge 2 commits into
Conversation
leaf-agent
left a comment
leaf-agent
left a comment
There was a problem hiding this comment.
Two things about the A/B as written, both about it returning "only the hash moved" when the gallery has actually moved — the outcome the paragraph exists to prevent.
The substitution that makes the A/B work also blinds it to one input class. theme.css is a capture input (capture_input_files() rglobs skills/leaf/assets), so editing --serif — reordering it, dropping Georgia, adding a face — reddens the test. On the runner every name in that stack resolves to the same substitute (fc-match Charter and fc-match "Iowan Old Style" both return DejaVu Sans here, fc-match Georgia returns DejaVu Serif), so both trees render identically and the A/B says hash-only, while on the authoring machine, where Charter exists, the gallery really did move. Worth the clause, since the font stack is exactly the input a reader of this paragraph is most likely to have touched.
The generator writes docs/example-*.jpg and rewrites the manifest in whichever tree it runs in. "again at the tip" reads as the session's own checkout, which leaves nine runner-font captures and a regenerated manifest as working-tree modifications right before the session is told to make a commit — and the regenerated manifest carries the runner captures' sha256 values, so committing it wholesale trades one red assertion for another. Naming both runs as throwaway worktrees closes that.
Separately, and not something I'd hold the change for: the paragraph names the baseline commit by property but not by how to find it, so the next session re-derives the search — computing digest(capture_input_files()) at candidate trees until one matches its manifest. A clause pointing at the manifest's own history as the candidate list would save that.
The measurements hold up. On the merged tree the manifest reads ff92edc6… against a computed 1936c559…, matching the body's tip row exactly.
Both generator runs go in throwaway worktrees, so the runner-font stills and the manifest they rewrite never land on the branch. The all-identical verdict now carries the theme.css font-stack exception, and the baseline commit is found from the manifest's own history rather than re-derived.
leaf-agent
left a comment
leaf-agent
left a comment
There was a problem hiding this comment.
main moved past this change while the last commit was being written. 4b42d15 "Remove the catalog preview freshness digest" landed at 22:51:04Z, five minutes before 7079fda, and it deletes exactly the four lines this PR rewrites — along with everything the new eighteen lines describe: docs/example-previews.json, capture_input_files()/digest()/inputs_sha256 in scripts/example-previews.py, and the assertion in tests/test_site.py that raised preview inputs changed — rerun scripts/example-previews.py. So the bullet is now keyed on a failure message nothing can emit, and prescribes an A/B over a manifest field that no longer exists. gh pr view 175 --json mergeable reports CONFLICTING for the same reason: the base lines are gone.
The three things the earlier review raised are all handled in 7079fda — throwaway worktrees, the font-stack blind spot, the baseline lookup — so this isn't a quality objection. There is just no longer a class to document: the maintainer's answer to "the digest can't tell CI whether the gallery moved" was to stop coupling the two, which is a level below where this PR was working. Recommend closing.
One measurement in the PR body outlives the mechanism. Six of the nine stills moved at bba885a, and examples/release-notes.html on main now reads "A filled margin Button commits its verb; a hollow one reveals context, and … unfolds more Buttons beside it." while the checked-in still predates that edit — I confirmed the sentence against 4b42d15; the capture dates are this PR’s own A/B. The digest was what flagged that drift, and #133 — which tracked it — closed earlier today, so with the digest gone nothing is watching: https://leaf.page/ is serving a preview of a sentence the example no longer has, and only the authoring machine has the faces to recapture it. Worth a home somewhere other than a PR that closes.
|
Closing: the mechanism this paragraph documents is gone from One verified fact outlives the mechanism, recorded here rather than as an issue since only the authoring machine can act on it: https://leaf.page/ serves a release-notes preview of a sentence the example no longer has. |
## Problem [`ci` run 33516407327](https://github.com/max-sixty/leaf/actions/runs/33516407327) came back **10 failed, 1614 passed, 6 skipped**. Eight belong to #189, whose `ci` run was cancelled by #190's push before the nightly leg ever ran on it. #188's `test_mcp_app.py` case arrived on the same collision as the MCP test below and is carried here rather than left to red the next nightly. The everyday suite that gates a pull request does not run `test_render_*` or `test_site.py`, so a PR can land its nightly half unread — which is how these arrived, and how five more have arrived since, from #193, #194 and #200. `main` has moved a long way under this branch; the ones it now owns are taken on merge, and three of the new ones are cleared here. #203 reached `tests/conftest.py`'s `headless_shell` first, so the four `test_render_commands.py` setup errors are its. ## Solution Grouped by cause rather than by test, since several tests share one. **The example the tests still name.** #189 rewrote `examples/pr-walkthrough.html` into the Worktrunk review packet, and the shipped patch is now a collapsed manifest that builds no lines until a reader opens a file, which the scroller gate read as `getComputedStyle(null)`. The gate reads the diffs that have drawn lines, with its own non-vacuity floor kept. **The new package's chrome and surfaces.** `.lf-pr-description` stacked a second translucent accent onto a card that is already a tint of one, and the layer's own `--muted` reads 4.32:1 there; the description now sits on the raised-surface token. CallDiff's disclosure was a bare `<button>` rather than an `offer`, so an exported copy kept a hand over a press nothing can take. Its location anchor carried an `href` on the header row, which names no location at all — and `reachScrollers` reads a candidate for a focusable descendant before granting it a stop, so that hidden anchor answered "there is already a way in here" with a link nobody can reach, for a box whose words run off the side. **A reading that counted clipped text as painted.** `coveredWords` measured each run's whole rect, so an ellipsised name in a narrow column read as covering its neighbour. It now intersects each run with the boxes that clip it, stopping at an out-of-flow ancestor where a hidden overflow further out need not reach. The one test that plants this fault plants it properly now: the runtime's note is parked in a one-pixel box with hidden overflow, so opacity alone never put its characters on the screen. **Three focus rings the corpus stopped painting.** `code-pre-light` and `code-pre-shadow` are the tab stops on a scrolling code block and on a diff's lines, and with the old example gone no example held either. The page gets one rendered hunk and one code line long enough to scroll. Then `.lf-diff-review`: it took the outward ring it shared with `.lf-diff-next`, which stands on its own in the toolbar, while this one rides a file's summary row inside a box that clips at the row's bottom edge — its lower run fell 3px past what the sweep can see. It takes the inset the summary beside it already takes, and splitting the rule made each half declare the ring name neither had. **`--diffs-fg-number` on changed rows.** Surfaced by that rendered hunk: a changed row's number is drawn in the base its own fill is mixed from, green on green at 4.03:1. Pierre's number overrides now take the layer's tinted-surface inks. **A sidebar the page got shorter under.** #190's zero-height ToC anchor shortened the release page 58px, putting a named `scrollTo(0, 900)` past the stretch where the box stands on its own offset. Both edges of that stretch are the page's, so they are read off it, with a floor asserting the stretch exists before a point halfway along it says anything. **Playwright's loop against `asyncio.run` and `anyio.run`.** `sync_playwright()` keeps an asyncio loop running in its thread for its whole lifetime, and the `browser` fixture is session-scoped per xdist worker — so an MCP test's own loop start raises in any worker that has opened a browser and passes in one that has not, leaving the schedule to decide. Verified directly: `asyncio.events._get_running_loop()` is `None` before `sync_playwright()`, the loop inside it, `None` after. `interact_support.run_async` runs the entry point on a thread with no loop on it, and both MCP modules go through it — #204's three new exchanges included, since they landed on the direct call while this was in flight. **A declared verb no page replayed.** #194 added `lf-diff`'s `review` to the registry without an event for it on the standing-state page — the one fixture built so that exactly this fails rather than going unexercised. It gets a diff and a standing review. **A focus contract two site tests never heard about.** #193 kept the response field passive so a drag leaves the browser's own selection alone and a native copy still has something to take, and aligned the render tests with it. `test_site.py` is nightly-only. Its label test turned on `document.activeElement === field`, which is now false for the page's own words and the site's label alike — the discriminator was gone, not inverted — so it reads whether the field was offered at all, which is the difference it was always after. ## Testing `uv run pytest tests --run-nightly` and `pre-commit run --all-files` on this head. The everyday suite and lint are green on this branch in CI. Every claim about what `main` does was taken from a control run of the same tests in a clean checkout of `main`, not from reading. <details><summary>Where each failure landed</summary> | test | fix | | --- | --- | | `test_render_gate.py::test_the_runtime_holds_a_scroller_the_page_wrote` | reads the diffs that drew lines | | `test_render_export.py::test_an_exported_example_stands_on_its_own[pr-walkthrough]`, `[corpus]` | widget, probe, theme | | `test_render_controls.py::test_every_ring_the_layer_draws_is_shown_whole_somewhere_in_the_corpus` | example, then `.lf-diff-review`'s ring and both ring names | | `test_render_pages.py::test_a_left_sidebar_uses_the_margin_until_the_page_needs_it_back` | scroll position read off the page | | `test_interact_mcp.py::test_stdio_protocol_carries_the_app_resource_and_private_tool_result` | own thread | | `test_mcp_app.py::test_registered_server_prefers_full_page_and_keeps_snapshot_as_fallback` | own thread | | `test_render_projection.py::test_the_render_gate_applies_every_standing_action_a_second_time` | standing fixture (#194's) | | `test_site.py::test_the_label_is_chrome_rather_than_words_to_quote`, `::test_a_comment_lands_in_the_thread_with_its_quote` | the field is offered, not entered (#193's) | | `test_render_anchors.py::test_a_data_bound_diff_aims_and_selects_one_source_line` | **#194's and #193's, taken on merge** | | `test_site.py`, `test_render_export.py` stale `<h1>` | **#202's, taken on merge** | The two exported-copy cases each carried four faults, three of them masked: `offering` asserts before `covered`, which asserts before the copy's axe run, so each fix uncovered the next. The ring sweep did the same — the geometry fault stood in front of a naming check that had been silent about a rule missing its name since it was written. </details> <details><summary>Three this leaves, all of them main's</summary> **`test_render_controls.py::test_examples_have_no_serious_wcag_a_or_aa_violations[pr-walkthrough]` stays red.** The `.lf-pr-description` contrast above is fixed and confirmed gone. What holds the test red is 46 `nested-interactive` findings from #194: each file's "Mark reviewed" `<button>` sits inside that file's `<summary>`, and a disclosure with a focusable descendant is a serious WCAG failure. Clearing it means taking the press out of the disclosure and re-laying the row — `entry.node` is the `<details>` in ten places, and the theme, print and export rules key on it — which is a change to that widget's shape with visual verification of its own, not a line this branch should slip in. **`test_render_anchors.py::test_a_failed_fragment_hydration_waits_for_a_reader_retry` fails whenever anything runs before it.** Confirmed on unmodified `main`: three runs of the file at `-n4`, three failures, and the same failure at `-n0` over the whole file. It passes alone every time. `assert len(requests) == 2` sees one — the reopened disclosure's second fragment fetch never reaches the route. Arrived with #189. **`test_render_anchors.py::test_a_drag_released_mid_word_hugs_words_and_sentences` fails over the whole file and passes alone.** `assert 'inside' == 'inside it'` — the selection did not grow to the word boundary. It looks like the next instance of what #191 diagnosed, Chromium 151 collapsing a synthetic range on mouse release. Also still behind: the gallery's `docs/example-pr-walkthrough.jpg`. Regenerating here rewrites all nine stills in substituted faces, because `theme.css` asks for Charter and this runner's `fc-match` has none of the serif stack — the same reading #175 made. Left for the authoring machine. </details> --- Automated fix for [failed run](https://github.com/max-sixty/leaf/actions/runs/33516407327) > _This was written by Claude Code on behalf of @max-sixty_ --------- Co-authored-by: leaf-agent <318509791+leaf-agent@users.noreply.github.com> Co-authored-by: Maximilian Roos <m@maxroos.com> Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
1 participant
Problem
cirun 33443479562 on0eec71acame back 3 failed, 1587 passed, 6 skipped. Two are the anchor pair #174 already carries and fixes; the third istest_site.py::test_the_public_catalog_is_a_visual_index_of_full_page_routes— the catalog preview digest, whose handling #168 had just named as its own red-suite class.I followed that class from CI and it sends a run somewhere it cannot go. "Run the generator and commit a manifest-only refresh" only holds when the stills did not move, and the generator run here cannot tell you that:
theme.cssasks forCharter, "Bitstream Charter", "Iowan Old Style", Georgia, ui-serif, serifandsystem-uifor sans,fc-matchon the runner has none of the first four, so every capture taken here differs from every checked-in still whether or not the tree changed. A run that follows the sentence literally either commits runner-font stills or bumpsinputs_sha256on faith.Solution
The class now says which of the two readings it is, and how to get it. Compare the generator's output to itself across two trees — a worktree at the last commit whose manifest equalled its computed digest, and one at the tip — rather than to what is checked in. The substituted faces cancel, so nine identical stills mean only the hash moved and an
inputs_sha256-only commit is the whole fix, while any difference means the gallery is genuinely behind and only the authoring machine has the faces to recapture it.Documentation only; no test or runtime change.
Testing
Ran the A/B for the live gap, which is what the wording is drawn from.
b932acbis the last commit onmainwhose checked-ininputs_sha256equalsdigest(capture_input_files()). Capturing there and atb38c6e0, same runner and Chrome: six of the nine stills moved, so this gap lands on the branch that must not be committed from here.b932acb→b38c6e0command-hubrelease-notesdesign-decision,heat-loss,live-progress,ship-reviewpostmortem,pr-walkthrough,triage-boardAll six moved at
bba885a"Canonicalize target Buttons" and nothing moved after it —0eec71a,bdf9f44andb38c6e0each moveinputs_sha256while leaving all nine captures byte-identical. The drift is a content change rather than a rendering artefact:examples/release-notes.html's first paragraph gained "A filled margin Button commits its verb; a hollow one reveals context, and … unfolds more Buttons beside it." and lost "decide the proposed replacement, deletion, and addition in the margin", so https://leaf.page/ is serving a sentence the example no longer has.pre-commit run --files .claude/skills/running-tend/SKILL.md: pass.What this PR does not do: clear the red. The digest on
main's tip78fa4c4computes1936c559f559e6a710e084dd24c45377bf6ce6d108bd6a032383eb244a46229fagainst a manifest still readingff92edc6…, and the six moved stills mean clearing it needsscripts/example-previews.pyre-run on the authoring machine, after the branch's own last merge ofmain. That is the same conclusion as #133, now with the measurement that says a hash-only bump would have been wrong this time — unlike the03bfc0d→7f58225gap, where all nine came back identical and it would have been right.How the A/B was run
scripts/example-previews.pyrun unmodified in a detached worktree per commit on theubuntu-24.04runner, same Chrome throughout, comparingsha256of the ninedocs/example-*.jpgbetween output sets. Trees:b932acb,bba885a,0eec71a,bdf9f44,b38c6e0; per-still change attributed by comparing consecutive pairs. Horizontal offsets are the shift minimising mean absolute difference over the page area below the banner, at the 896×560 output scale, so ×1.25 for viewport pixels.digest(capture_input_files())per tree, against the checked-ininputs_sha256read withgit show <commit>:docs/example-previews.json:b932acbff92edc6…ff92edc6…mainbba885aff92edc6…012f7429…0eec71aff92edc6…09b2364f…bdf9f44ff92edc6…df040ed3…b38c6e0ff92edc6…c493c61f…78fa4c4(tip)ff92edc6…1936c559…The
0eec71arow reproducing the run's assertion is what says the method matches the gate's.Fonts, measured rather than recalled —
fc-liston the runner holds DejaVu, Lato, Liberation and Noto Color Emoji only;fc-match Charter,fc-match "Bitstream Charter"andfc-match "Iowan Old Style"all return DejaVu Sans,fc-match Georgiareturns DejaVu Serif.Pixel counts for the
b932acb→b38c6e0pair, threshold 12/255:command-hub11.4%,design-decision10.5%,heat-loss9.5%,ship-review8.9%,release-notes7.4%,live-progress6.2%, and exactly 0 for the other three.Automated fix for failed run