diff --git a/.aegis-baseline.tsv b/.aegis-baseline.tsv index 746be585..97ec6a73 100644 --- a/.aegis-baseline.tsv +++ b/.aegis-baseline.tsv @@ -24,24 +24,42 @@ # points, and the GitHub release, asset, run and check identifiers # quoted in the release receipts; this repository holds no tax # identifiers. +australian-tfn ./docs/TASKS.md 90854b2dc477585b494af511ebe7e260c09e9e6d1582f8d75c35733426d6d710 2 +australian-tfn ./docs/release/ARTIFACT-MATRIX.md 90854b2dc477585b494af511ebe7e260c09e9e6d1582f8d75c35733426d6d710 9 +australian-tfn ./docs/release/CERTIFICATION.md 90854b2dc477585b494af511ebe7e260c09e9e6d1582f8d75c35733426d6d710 2 australian-tfn ./docs/release/EXECUTION-BOARD.md 35b904df2c6f31c493260dde331b4104ae508d471dc06ce7595bf3ad5a681134 3 -australian-tfn ./docs/release/EXECUTION-BOARD.md a3a2cc550be9d1627cd6ebe8554832486df9ae30304ab8ca9b141b1f88bb4e99 2 +australian-tfn ./docs/release/EXECUTION-BOARD.md 90854b2dc477585b494af511ebe7e260c09e9e6d1582f8d75c35733426d6d710 1 +australian-tfn ./docs/release/EXECUTION-BOARD.md a3a2cc550be9d1627cd6ebe8554832486df9ae30304ab8ca9b141b1f88bb4e99 1 +australian-tfn ./docs/release/EXECUTION-BOARD.md f56a84c068b5f3241a93c04e12a698d51e771dfef255452d1b225b52854eaf76 1 +australian-tfn ./docs/release/LINUX-COMPAT.md f56a84c068b5f3241a93c04e12a698d51e771dfef255452d1b225b52854eaf76 1 australian-tfn ./docs/release/RECEIPT-0.2.5-DRYRUN-3.md 35b904df2c6f31c493260dde331b4104ae508d471dc06ce7595bf3ad5a681134 1 australian-tfn ./docs/release/RECEIPT-0.2.5-DRYRUN-3.md a3a2cc550be9d1627cd6ebe8554832486df9ae30304ab8ca9b141b1f88bb4e99 1 +australian-tfn ./docs/release/RECEIPT-0.2.5-DRYRUN-6.md 90854b2dc477585b494af511ebe7e260c09e9e6d1582f8d75c35733426d6d710 2 +australian-tfn ./docs/release/RECEIPT-0.2.5-DRYRUN-6.md f56a84c068b5f3241a93c04e12a698d51e771dfef255452d1b225b52854eaf76 1 australian-tfn ./docs/research/fireworks-png-nodes.md 1004a585bc60d363038c5889d022bd6f58278338fe66f97483316133e316f41a 1 australian-tfn ./docs/tasks/T27-19-draft-release-test-tag.md 35b904df2c6f31c493260dde331b4104ae508d471dc06ce7595bf3ad5a681134 1 -australian-tfn ./docs/tasks/T27-52-source-archive.md a3a2cc550be9d1627cd6ebe8554832486df9ae30304ab8ca9b141b1f88bb4e99 1 +australian-tfn ./docs/tasks/T27-19-draft-release-test-tag.md 90854b2dc477585b494af511ebe7e260c09e9e6d1582f8d75c35733426d6d710 1 +australian-tfn ./docs/tasks/T27-19-draft-release-test-tag.md f56a84c068b5f3241a93c04e12a698d51e771dfef255452d1b225b52854eaf76 1 australian-tfn ./xtask/src/aegis.rs 1004a585bc60d363038c5889d022bd6f58278338fe66f97483316133e316f41a 1 # Matches the same 9-digit observed value range in the PNG node # dictionary and the GitHub release, asset, run and check identifiers # quoted in the release receipts; this repository holds no banking data. +bank-routing-number ./docs/TASKS.md 90854b2dc477585b494af511ebe7e260c09e9e6d1582f8d75c35733426d6d710 2 +bank-routing-number ./docs/release/ARTIFACT-MATRIX.md 90854b2dc477585b494af511ebe7e260c09e9e6d1582f8d75c35733426d6d710 9 +bank-routing-number ./docs/release/CERTIFICATION.md 90854b2dc477585b494af511ebe7e260c09e9e6d1582f8d75c35733426d6d710 2 bank-routing-number ./docs/release/EXECUTION-BOARD.md 35b904df2c6f31c493260dde331b4104ae508d471dc06ce7595bf3ad5a681134 3 -bank-routing-number ./docs/release/EXECUTION-BOARD.md a3a2cc550be9d1627cd6ebe8554832486df9ae30304ab8ca9b141b1f88bb4e99 2 +bank-routing-number ./docs/release/EXECUTION-BOARD.md 90854b2dc477585b494af511ebe7e260c09e9e6d1582f8d75c35733426d6d710 1 +bank-routing-number ./docs/release/EXECUTION-BOARD.md a3a2cc550be9d1627cd6ebe8554832486df9ae30304ab8ca9b141b1f88bb4e99 1 +bank-routing-number ./docs/release/EXECUTION-BOARD.md f56a84c068b5f3241a93c04e12a698d51e771dfef255452d1b225b52854eaf76 1 +bank-routing-number ./docs/release/LINUX-COMPAT.md f56a84c068b5f3241a93c04e12a698d51e771dfef255452d1b225b52854eaf76 1 bank-routing-number ./docs/release/RECEIPT-0.2.5-DRYRUN-3.md 35b904df2c6f31c493260dde331b4104ae508d471dc06ce7595bf3ad5a681134 1 bank-routing-number ./docs/release/RECEIPT-0.2.5-DRYRUN-3.md a3a2cc550be9d1627cd6ebe8554832486df9ae30304ab8ca9b141b1f88bb4e99 1 +bank-routing-number ./docs/release/RECEIPT-0.2.5-DRYRUN-6.md 90854b2dc477585b494af511ebe7e260c09e9e6d1582f8d75c35733426d6d710 2 +bank-routing-number ./docs/release/RECEIPT-0.2.5-DRYRUN-6.md f56a84c068b5f3241a93c04e12a698d51e771dfef255452d1b225b52854eaf76 1 bank-routing-number ./docs/research/fireworks-png-nodes.md 1004a585bc60d363038c5889d022bd6f58278338fe66f97483316133e316f41a 1 bank-routing-number ./docs/tasks/T27-19-draft-release-test-tag.md 35b904df2c6f31c493260dde331b4104ae508d471dc06ce7595bf3ad5a681134 1 -bank-routing-number ./docs/tasks/T27-52-source-archive.md a3a2cc550be9d1627cd6ebe8554832486df9ae30304ab8ca9b141b1f88bb4e99 1 +bank-routing-number ./docs/tasks/T27-19-draft-release-test-tag.md 90854b2dc477585b494af511ebe7e260c09e9e6d1582f8d75c35733426d6d710 1 +bank-routing-number ./docs/tasks/T27-19-draft-release-test-tag.md f56a84c068b5f3241a93c04e12a698d51e771dfef255452d1b225b52854eaf76 1 bank-routing-number ./xtask/src/aegis.rs 1004a585bc60d363038c5889d022bd6f58278338fe66f97483316133e316f41a 1 # The GitForge API row names the credential file that holds the bearer # token; the row contains no token. @@ -209,13 +227,22 @@ sql-query ./docs/tasks/T22-screenshot-ui.md 6ba454ee08d4ce45364f6e849176079be5c6 # dictionary and the GitHub release, asset, run and check identifiers # quoted in the release receipts; this repository holds no # social-security numbers. +ssn-no-dashes ./docs/TASKS.md 90854b2dc477585b494af511ebe7e260c09e9e6d1582f8d75c35733426d6d710 2 +ssn-no-dashes ./docs/release/ARTIFACT-MATRIX.md 90854b2dc477585b494af511ebe7e260c09e9e6d1582f8d75c35733426d6d710 9 +ssn-no-dashes ./docs/release/CERTIFICATION.md 90854b2dc477585b494af511ebe7e260c09e9e6d1582f8d75c35733426d6d710 2 ssn-no-dashes ./docs/release/EXECUTION-BOARD.md 35b904df2c6f31c493260dde331b4104ae508d471dc06ce7595bf3ad5a681134 3 -ssn-no-dashes ./docs/release/EXECUTION-BOARD.md a3a2cc550be9d1627cd6ebe8554832486df9ae30304ab8ca9b141b1f88bb4e99 2 +ssn-no-dashes ./docs/release/EXECUTION-BOARD.md 90854b2dc477585b494af511ebe7e260c09e9e6d1582f8d75c35733426d6d710 1 +ssn-no-dashes ./docs/release/EXECUTION-BOARD.md a3a2cc550be9d1627cd6ebe8554832486df9ae30304ab8ca9b141b1f88bb4e99 1 +ssn-no-dashes ./docs/release/EXECUTION-BOARD.md f56a84c068b5f3241a93c04e12a698d51e771dfef255452d1b225b52854eaf76 1 +ssn-no-dashes ./docs/release/LINUX-COMPAT.md f56a84c068b5f3241a93c04e12a698d51e771dfef255452d1b225b52854eaf76 1 ssn-no-dashes ./docs/release/RECEIPT-0.2.5-DRYRUN-3.md 35b904df2c6f31c493260dde331b4104ae508d471dc06ce7595bf3ad5a681134 1 ssn-no-dashes ./docs/release/RECEIPT-0.2.5-DRYRUN-3.md a3a2cc550be9d1627cd6ebe8554832486df9ae30304ab8ca9b141b1f88bb4e99 1 +ssn-no-dashes ./docs/release/RECEIPT-0.2.5-DRYRUN-6.md 90854b2dc477585b494af511ebe7e260c09e9e6d1582f8d75c35733426d6d710 2 +ssn-no-dashes ./docs/release/RECEIPT-0.2.5-DRYRUN-6.md f56a84c068b5f3241a93c04e12a698d51e771dfef255452d1b225b52854eaf76 1 ssn-no-dashes ./docs/research/fireworks-png-nodes.md 1004a585bc60d363038c5889d022bd6f58278338fe66f97483316133e316f41a 1 ssn-no-dashes ./docs/tasks/T27-19-draft-release-test-tag.md 35b904df2c6f31c493260dde331b4104ae508d471dc06ce7595bf3ad5a681134 1 -ssn-no-dashes ./docs/tasks/T27-52-source-archive.md a3a2cc550be9d1627cd6ebe8554832486df9ae30304ab8ca9b141b1f88bb4e99 1 +ssn-no-dashes ./docs/tasks/T27-19-draft-release-test-tag.md 90854b2dc477585b494af511ebe7e260c09e9e6d1582f8d75c35733426d6d710 1 +ssn-no-dashes ./docs/tasks/T27-19-draft-release-test-tag.md f56a84c068b5f3241a93c04e12a698d51e771dfef255452d1b225b52854eaf76 1 ssn-no-dashes ./xtask/src/aegis.rs 1004a585bc60d363038c5889d022bd6f58278338fe66f97483316133e316f41a 1 # The rule's regex fires on the HTTP client the Aegis installer uses — # the node `fetch(url)` fallback in `scripts/install-aegis.sh` — and diff --git a/.claude/scratch/run_engine_test.sh b/.claude/scratch/run_engine_test.sh deleted file mode 100644 index d7960994..00000000 --- a/.claude/scratch/run_engine_test.sh +++ /dev/null @@ -1,23 +0,0 @@ -#!/usr/bin/env bash -# Wait for local cargo admission (no competing compiler processes and -# host PSI under the gate), then run the engine history-rewind test. -set -u -cd /nas/Temp/repos/FireCraft/.claude/worktrees/wt-hist-session || exit 1 - -while true; do - compilers=$(ps -eo args | grep -E '(cargo-clippy|/cargo (build|check|clippy|test|run)|/rustc --crate-name)' | grep -v 'grep -E' | wc -l) - cpu_some=$(awk '$1=="some"{print $2}' /proc/pressure/cpu | sed 's/avg10=//;s/%//') - io_some=$(awk '$1=="some"{print $2}' /proc/pressure/io | sed 's/avg10=//;s/%//') - io_full=$(awk '$1=="full"{print $2}' /proc/pressure/io | sed 's/avg10=//;s/%//') - if [ "$compilers" -eq 0 ] && \ - awk -v c="$cpu_some" 'BEGIN{exit !(c<20)}' && \ - awk -v s="$io_some" 'BEGIN{exit !(s<20)}' && \ - awk -v f="$io_full" 'BEGIN{exit !(f<10)}'; then - break - fi - sleep 20 -done - -echo "admission open; running the engine test" -CARGO_TARGET_DIR=/var/tmp/fc-rewind-target cargo test -p engine --test history_rewind 2>&1 -echo "exit=$?" diff --git a/.github/workflows/release.yml b/.github/workflows/release.yml index 783aa389..e153180d 100644 --- a/.github/workflows/release.yml +++ b/.github/workflows/release.yml @@ -66,6 +66,13 @@ jobs: # the tree), and on a rehearsal the tag is a disposable name # while the artifacts still describe the version they carry. id: version + env: + # The inputs come through the environment, not through + # interpolation into the script: a rehearsal tag is a string a + # human typed, and it must never become part of the shell. + EVENT_NAME: ${{ github.event_name }} + REHEARSAL_TAG_INPUT: ${{ inputs.rehearsal_tag }} + REF_NAME: ${{ github.ref_name }} run: | set -euo pipefail version=$(cargo metadata --no-deps --format-version 1 \ @@ -75,7 +82,17 @@ jobs: exit 1 fi echo "version=$version" >> "$GITHUB_OUTPUT" - echo "the workspace builds version $version" + # The tag the release will carry: the pushed tag on a tag + # build, the disposable rehearsal tag on a manual one. It is + # stated, not inferred — a rehearsal tag does not exist yet + # when the artifacts are built. + if [ "$EVENT_NAME" = "workflow_dispatch" ]; then + tag="$REHEARSAL_TAG_INPUT" + else + tag="$REF_NAME" + fi + echo "release_tag=$tag" >> "$GITHUB_OUTPUT" + echo "the workspace builds version $version for tag $tag" - name: Install Rust uses: dtolnay/rust-toolchain@b725eda525362e7cf34af9a7243cac1e3f01a7ec # stable @@ -209,6 +226,14 @@ jobs: # timestamp of the run that produced it. It is generated after # every asset exists, and the verifier then re-reads it against # the files on disk. + # + # The tag is *stated*, not inferred: a rehearsal tag does not + # exist while the artifacts are built, so neither the git tag nor + # the CI ref name can name it. xtask validates the stated tag + # against the workspace version and refuses to write a manifest + # whose tag and tree disagree. + env: + RELEASE_TAG: ${{ steps.version.outputs.release_tag }} run: | cargo run --quiet --package xtask -- dist --manifest cargo run --quiet --package xtask -- dist --verify-manifest diff --git a/Cargo.lock b/Cargo.lock index fef5d566..99e60d01 100644 --- a/Cargo.lock +++ b/Cargo.lock @@ -201,6 +201,7 @@ dependencies = [ "eframe", "egui", "egui_extras", + "egui_kittest", "engine", "flate2", "fwpng", @@ -1339,6 +1340,19 @@ dependencies = [ "winit", ] +[[package]] +name = "egui_kittest" +version = "0.36.2" +source = "registry+https://github.com/rust-lang/crates.io-index" +checksum = "5eb3f3effb2726fee23449293e67acbf6cc6ce35525b1722bd5743bb348114d5" +dependencies = [ + "egui", + "kittest", + "log", + "serde", + "toml", +] + [[package]] name = "either" version = "1.19.0" @@ -2390,6 +2404,16 @@ version = "3.1.0" source = "registry+https://github.com/rust-lang/crates.io-index" checksum = "e2db585e1d738fc771bf08a151420d3ed193d9d895a36df7f6f8a9456b911ddc" +[[package]] +name = "kittest" +version = "0.4.0" +source = "registry+https://github.com/rust-lang/crates.io-index" +checksum = "90ceaa75eb0036a32b6b9833962eb18137449e9817e2e586006471925b727fd5" +dependencies = [ + "accesskit", + "accesskit_consumer 0.35.0", +] + [[package]] name = "kurbo" version = "0.11.3" @@ -3843,6 +3867,15 @@ dependencies = [ "syn 3.0.7", ] +[[package]] +name = "serde_spanned" +version = "1.1.3" +source = "registry+https://github.com/rust-lang/crates.io-index" +checksum = "9c187e0a369471c550af028cba56d4f4fa2a489ad28151ae95a4d340a2a8b782" +dependencies = [ + "serde_core", +] + [[package]] name = "sha2" version = "0.10.9" @@ -4340,11 +4373,24 @@ dependencies = [ "pin-project-lite", ] +[[package]] +name = "toml" +version = "1.2.1+spec-1.1.0" +source = "registry+https://github.com/rust-lang/crates.io-index" +checksum = "d5755c51863cdac711f1956baf8eb9588e060755f7d3c4c979e70ea1e1f32cd7" +dependencies = [ + "serde_core", + "serde_spanned", + "toml_datetime", + "toml_parser", + "winnow 1.0.4", +] + [[package]] name = "toml_datetime" -version = "1.1.2+spec-1.1.0" +version = "1.1.3+spec-1.1.0" source = "registry+https://github.com/rust-lang/crates.io-index" -checksum = "2b86d767906c6c42421dcba507eb9d203e779497710a47782a224bb871653053" +checksum = "173f25c883cd2176a2d8212155638c0a60d484a9d84546d270866f40f56b6e49" dependencies = [ "serde_core", ] diff --git a/crates/app-cli/tests/web_behavior_export.rs b/crates/app-cli/tests/web_behavior_export.rs new file mode 100644 index 00000000..4d14db81 --- /dev/null +++ b/crates/app-cli/tests/web_behavior_export.rs @@ -0,0 +1,190 @@ +//! The web export's behavior half, end to end: a behavior attached +//! through the engine's command surface reaches the exported HTML as +//! the MM_* markup, and the file it swaps to is one the export +//! actually wrote. +//! +//! The exporter test (`webexport`) pins the markup, but it builds the +//! behavior on the model. This drives the same markup from the +//! command surface the shell drives — `slice.create`, +//! `selection.set`, `slice.addSwapImageBehaviorFromPoint`, +//! `slice.addBehavior`, `slice.setData` — through the export door the +//! CLI verb and the Export dialog share (`app_cli::ops::export_web`). + +#![allow(clippy::unwrap_used, clippy::expect_used, clippy::panic)] + +use app_cli::ops::export_web; +use engine::CommandRegistry; +use engine::catalog; +use webexport::{WebExportOptions, Workflow}; + +/// The engine's command registry with the built-in catalog installed. +fn registry() -> CommandRegistry { + let mut reg = CommandRegistry::new(); + let _ = catalog::install(&mut reg); + reg +} + +/// An 8x4 canvas of ink — red on the left, blue on the right — as +/// the raster import path builds it. +fn fixture() -> doc::Document { + let (width, height) = (8u32, 4u32); + let mut rgba = Vec::with_capacity(width as usize * height as usize * 4); + for _ in 0..height { + for x in 0..width { + let pixel = if x < 4 { + [255u8, 0, 0, 255] + } else { + [0u8, 0, 0, 255] + }; + rgba.extend_from_slice(&pixel); + } + } + io::raster::document_from_image( + &io::raster::DecodedImage { + width, + height, + rgba, + source: io::raster::SourceFormat::Png, + }, + Some("behaviors".to_owned()), + ) + .expect("fixture document") +} + +/// The id of the slice the document carries under `name`. +fn slice_id(document: &doc::Document, name: &str) -> doc::ObjectId { + document + .registry + .iter() + .find_map(|(id, data)| match data { + doc::ObjectData::Slice(slice) if slice.name.as_deref() == Some(name) => Some(id), + _ => None, + }) + .expect("the named slice") +} + +/// The payload form of an object id. +fn id_json(id: doc::ObjectId) -> serde_json::Value { + serde_json::json!({ "index": id.index(), "generation": id.generation() }) +} + +/// Two slices on the page's Web Layer, created through the command +/// surface (which provisions the layer itself). +fn sliced(registry: &CommandRegistry, document: &mut doc::Document) { + for (name, x) in [("left", 0.0), ("right", 4.0)] { + registry + .execute( + document, + "slice.create", + serde_json::json!({ + "shape": "rect", "x": x, "y": 0.0, + "width": 4.0, "height": 4.0, "page": 0, "name": name, + }), + ) + .expect("slice.create"); + } +} + +#[test] +fn a_swap_behavior_attached_through_the_command_surface_reaches_the_html() { + let mut document = fixture(); + let registry = registry(); + sliced(®istry, &mut document); + + // CS6's "add swap image behavior from point" verb: the selected + // slice gains the behavior, the slice under the point is the + // swap's source. + let left = slice_id(&document, "left"); + registry + .execute( + &mut document, + "selection.set", + serde_json::json!({ "ids": [id_json(left)] }), + ) + .expect("selection.set"); + registry + .execute( + &mut document, + "slice.addSwapImageBehaviorFromPoint", + serde_json::json!({ "where": { "x": 6.0, "y": 2.0 } }), + ) + .expect("addSwapImageBehaviorFromPoint"); + + let exported = export_web( + &document, + &WebExportOptions::default(), + Workflow::HtmlAndImages, + ) + .expect("web export"); + let html = exported.html.as_deref().expect("the HTML document"); + + // The behavior's markup rides the export. + assert!(html.contains("function MM_swapImage()"), "{html}"); + assert!( + html.contains("onMouseOver=\"MM_swapImage('left','','right.png',1);\""), + "{html}" + ); + assert!( + html.contains("onMouseOut=\"MM_swapImgRestore();\""), + "{html}" + ); + // The target is the file the export wrote for the slice the + // behavior names — not the bare name, which no file carries. + assert!( + html.contains("onload=\"MM_preloadImages('right.png')\""), + "{html}" + ); + assert!( + exported.files.iter().any(|file| file.name == "right.png"), + "the swap target is among the exported files: {:?}", + exported + .files + .iter() + .map(|file| file.name.clone()) + .collect::>() + ); +} + +#[test] +fn a_typed_file_name_target_is_emitted_verbatim() { + let mut document = fixture(); + let registry = registry(); + sliced(®istry, &mut document); + + // The other command path: the behavior and its target written + // explicitly, with a file name the user typed. + let left = slice_id(&document, "left"); + registry + .execute( + &mut document, + "slice.addBehavior", + serde_json::json!({ "id": id_json(left), "kind": 2 }), + ) + .expect("slice.addBehavior"); + registry + .execute( + &mut document, + "slice.setData", + serde_json::json!({ + "id": id_json(left), "key": "swapTarget", "value": "hero.png", + }), + ) + .expect("slice.setData"); + + let exported = export_web( + &document, + &WebExportOptions::default(), + Workflow::HtmlAndImages, + ) + .expect("web export"); + let html = exported.html.as_deref().expect("the HTML document"); + + assert!( + html.contains("onMouseOver=\"MM_swapImage('left','','hero.png',1);\""), + "{html}" + ); + assert!( + html.contains("onload=\"MM_preloadImages('hero.png')\""), + "{html}" + ); +} diff --git a/crates/app/Cargo.toml b/crates/app/Cargo.toml index 03d96ebf..21fce514 100644 --- a/crates/app/Cargo.toml +++ b/crates/app/Cargo.toml @@ -44,6 +44,11 @@ scripting = { path = "../scripting" } # `ron` dep is private), so the layout round-trip test exercises the # exact on-disk shape. ron = "0.12" +# egui's own test harness: the shell's panels are drawn through it, +# so a panel's widgets are asserted as widgets rather than as pixels. +# The image-snapshot half needs a renderer (the `wgpu` feature) and +# is tracked separately (the ledger's queued item 21). +egui_kittest = "0.36" [lints] workspace = true diff --git a/crates/app/src/shell/html_preview.rs b/crates/app/src/shell/html_preview.rs index a6ddd215..51037b77 100644 --- a/crates/app/src/shell/html_preview.rs +++ b/crates/app/src/shell/html_preview.rs @@ -697,6 +697,38 @@ mod tests { assert!(cache.layout.width > 0.0); } + /// The preview's own widgets are asserted through egui's harness. + /// + /// The harness draws the panel into the accessibility tree a + /// screen reader reads, so the panel's labels are queried as + /// widgets rather than compared as pixels. The image-snapshot + /// half of this coverage needs a renderer (`egui_kittest`'s `wgpu` + /// feature) and a verified Vulkan adapter, which the CI image + /// does not carry yet — that half is the ledger's queued item 21. + #[test] + fn the_preview_draws_its_widgets_through_the_harness() { + use egui_kittest::kittest::Queryable as _; + + let mut shell = preview_shell("plate"); + let mut harness = egui_kittest::Harness::new_ui(move |ui| { + shell.html_preview_panel(ui); + }); + harness.run(); + + let source = harness.query_by_label_contains("HTML source"); + assert!( + source.is_some(), + "the preview offers the exported markup:\n{:#?}", + harness.root() + ); + let files = harness.query_by_label_contains("image file(s)"); + assert!( + files.is_some(), + "the preview reports the exported files:\n{:#?}", + harness.root() + ); + } + /// The Polygon Hotspot tool commits through the append state /// machine (decision 0029's Design C): point 1 takes /// `hotspot.appendPoint`'s create branch, each later point diff --git a/crates/app/src/shell/panels_web.rs b/crates/app/src/shell/panels_web.rs index 7a238e7a..0cb8699f 100644 --- a/crates/app/src/shell/panels_web.rs +++ b/crates/app/src/shell/panels_web.rs @@ -1711,8 +1711,13 @@ mod tests { "the exported page carries the behavior:\n{html}" ); assert!( - html.contains("'button'"), - "the exported page names the swap target:\n{html}" + html.contains("MM_swapImage('plate','','button.png',1);"), + "the exported page names the file the target exported as:\n{html}" + ); + assert!( + export.files.iter().any(|file| file.name == "button.png"), + "the file the swap points at is one the export wrote:\n{:#?}", + export.files ); } diff --git a/crates/webexport/src/html.rs b/crates/webexport/src/html.rs index c3fc64db..4c1b9418 100644 --- a/crates/webexport/src/html.rs +++ b/crates/webexport/src/html.rs @@ -4,6 +4,8 @@ //! emission for hotspots. HTML 4 variant, no comments (deterministic //! bytes). +use std::collections::BTreeMap; + use crate::images::Piece; use crate::layout::Grid; use crate::naming; @@ -24,17 +26,63 @@ fn escape(text: &str) -> String { out } +/// The file each named web object wrote, keyed by every name a +/// behavior can name it by: the slice's object name, its internal +/// name, and each of those cleaned. +/// +/// A behavior's `swapTarget` names the slice it points at — the +/// engine's `slice.addSwapImageBehaviorFromPoint` stores the source +/// slice's object name — while the file a swap points to is an +/// export-time artifact (`naming`), never stored on the slice +/// (decision 0017). The exporter therefore resolves the name to the +/// file it just wrote (decision 0133). A target that names no slice +/// is a file name the user typed and is emitted verbatim. +fn target_files(objects: &[WebObject<'_>], pieces: &[Piece]) -> BTreeMap { + let mut files: BTreeMap = BTreeMap::new(); + for piece in pieces { + let Some(index) = piece.object else { continue }; + let slice = objects[index].slice; + let mut names: Vec = Vec::new(); + for name in [slice.name.as_deref(), slice.internal_name.as_deref()] + .into_iter() + .flatten() + { + names.push(name.to_owned()); + names.push(naming::clean(name)); + } + for name in names { + if !name.is_empty() { + files.entry(name).or_insert_with(|| piece.file.clone()); + } + } + } + files +} + +/// The files the export wrote, and the directory it wrote them in. +/// +/// The two travel together: a behavior's target resolves to a file +/// the export wrote, and that file is referenced where it was +/// written, under the export's image directory (decision 0133). +struct Targets<'a> { + files: &'a BTreeMap, + images_dir: &'a str, +} + /// The behavior hooks one object's ``/`` carries, plus the /// files its behaviors preload: `(onMouseOver, onMouseOut, onClick, /// preloads)`. Kinds follow the template's constants — 2 swap image, /// 4 button down, 5 swap restore, 6 button highlight, 7 button -/// restore. The swap/down file rides `DTAv` custom data under -/// `swapTarget` (decision 0017); a behavior without a target emits no -/// hook. Popup menus (9) and unobserved kinds stay unrendered (M11 -/// non-goal): no hook, payload preserved in the file. +/// restore. The swap/down target rides `DTAv` custom data under +/// `swapTarget` (decision 0017) and is resolved to the file the +/// export wrote for the slice it names, at the path it wrote it +/// (decision 0133); a behavior without a target emits no hook. Popup +/// menus (9) and unobserved kinds stay unrendered (M11 non-goal): no +/// hook, payload preserved in the file. fn hooks( name: &str, slice: &doc::SliceObject, + targets: &Targets<'_>, ) -> (Option, Option, Option, Vec) { let mut over = String::new(); let mut out = String::new(); @@ -44,19 +92,27 @@ fn hooks( .custom_data .get("swapTarget") .map(String::as_str) - .filter(|target| !target.is_empty()); + .filter(|target| !target.is_empty()) + .map(|target| match targets.files.get(target) { + // A file the export wrote: reference it where it was + // written, next to the `` that names it. + Some(file) => format!("{}{file}", targets.images_dir), + // A file name the user typed: their own reference, used + // as given. + None => target.to_owned(), + }); for behavior in &slice.behaviors { match behavior.kind { // Nav-bar highlight: over. 6 => { - if let Some(target) = target { + if let Some(target) = target.as_deref() { over.push_str(&format!("MM_nbGroup('over','{name}','{target}','',1);")); preload.push(target.to_owned()); } } // Nav-bar down: click. 4 => { - if let Some(target) = target { + if let Some(target) = target.as_deref() { click.push_str(&format!( "MM_nbGroup('down','navbar1','{name}','{target}',1);" )); @@ -65,7 +121,7 @@ fn hooks( } // Swap image: over, restore on out. 2 => { - if let Some(target) = target { + if let Some(target) = target.as_deref() { over.push_str(&format!("MM_swapImage('{name}','','{target}',1);")); preload.push(target.to_owned()); out.push_str("MM_swapImgRestore();"); @@ -196,7 +252,12 @@ fn cut_index(cuts: &[u32], at: f64) -> Option { /// Anchors every hotspot (a `TSLi`-less web object) to the piece /// containing its center, in object order. -fn anchor_hotspots(objects: &[WebObject<'_>], grid: &Grid, pieces: &[Piece]) -> Vec { +fn anchor_hotspots( + objects: &[WebObject<'_>], + grid: &Grid, + pieces: &[Piece], + targets: &Targets<'_>, +) -> Vec { let mut anchors = Vec::new(); for (index, object) in objects.iter().enumerate() { if object.slice.slice_type.is_some() { @@ -222,7 +283,7 @@ fn anchor_hotspots(objects: &[WebObject<'_>], grid: &Grid, pieces: &[Piece]) -> // anchor name is the host image (the swap runtime addresses // images, and an area has none of its own). let host = naming::stem_of(&pieces[piece_index].file); - let (over, out, click, _) = hooks(host, object.slice); + let (over, out, click, _) = hooks(host, object.slice, targets); anchors.push(Anchor { piece: piece_index, hotspot: index, @@ -246,7 +307,12 @@ pub(crate) fn write_html( let Some(page) = document.pages.get(options.page) else { return String::new(); }; - let anchors = anchor_hotspots(objects, grid, pieces); + let files = target_files(objects, pieces); + let targets = Targets { + files: &files, + images_dir: &options.images_dir, + }; + let anchors = anchor_hotspots(objects, grid, pieces, &targets); let width = grid.x_cuts.last().copied().unwrap_or(0); // Behavior pass: runtime need and preload list, pieces first then @@ -268,7 +334,7 @@ pub(crate) fn write_html( if let Some(index) = piece.object { let name = naming::stem_of(&piece.file); note( - hooks(name, objects[index].slice), + hooks(name, objects[index].slice, &targets), &mut preloads, &mut needs_runtime, ); @@ -276,7 +342,11 @@ pub(crate) fn write_html( } for anchor in &anchors { let hotspot = objects[anchor.hotspot].slice; - note(hooks("area", hotspot), &mut preloads, &mut needs_runtime); + note( + hooks("area", hotspot, &targets), + &mut preloads, + &mut needs_runtime, + ); } let mut html = String::with_capacity(4096); @@ -342,12 +412,12 @@ pub(crate) fn write_html( next_col = piece.col + piece.col_span; write_piece( &mut html, - options, objects, grid, piece_index, piece, &anchors, + &targets, ); } if options.shim { @@ -408,12 +478,12 @@ pub(crate) fn write_html( /// piece's position in the rasterizer's output. fn write_piece( html: &mut String, - options: &WebExportOptions, objects: &[WebObject<'_>], grid: &Grid, piece_index: usize, piece: &Piece, anchors: &[Anchor], + targets: &Targets<'_>, ) { let (_, _, w, h) = grid.cell_rect(piece.row, piece.col); let slice = piece.object.map(|index| objects[index].slice); @@ -430,7 +500,7 @@ fn write_piece( let name = naming::stem_of(&piece.file); let mut anchor_open = false; if let Some(slice) = slice { - let (over, out, click, _) = hooks(name, slice); + let (over, out, click, _) = hooks(name, slice, targets); if over.is_some() || out.is_some() || click.is_some() || slice.link.is_some() { let href = slice .link @@ -451,7 +521,7 @@ fn write_piece( anchor_open = true; } } - let src = format!("{}{}", options.images_dir, piece.file); + let src = format!("{}{}", targets.images_dir, piece.file); let alt = slice .and_then(|slice| slice.alt.as_deref()) .map(escape) diff --git a/crates/webexport/tests/webexport.rs b/crates/webexport/tests/webexport.rs index c6d552e1..10fcacf9 100644 --- a/crates/webexport/tests/webexport.rs +++ b/crates/webexport/tests/webexport.rs @@ -251,6 +251,43 @@ fn behaviors_emit_hooks_and_the_runtime() { assert!(html.contains("onload=\"MM_preloadImages('swap.png')\"")); } +#[test] +fn a_behavior_naming_a_slice_swaps_to_the_file_that_slice_exported() { + let mut document = sliced_document(); + let left = document + .registry + .iter() + .find_map(|(id, data)| match data { + doc::ObjectData::Slice(slice) if slice.name.as_deref() == Some("left") => Some(id), + _ => None, + }) + .unwrap(); + let Some(doc::ObjectData::Slice(slice)) = document.registry.get_mut(left) else { + panic!("the left slice") + }; + // The behavior names the *slice* it points at; the file it swaps + // to is an export-time artifact (decision 0133). + slice.behaviors.push(doc::SliceBehavior { kind: 2 }); + slice + .custom_data + .insert("swapTarget".to_owned(), "right".to_owned()); + + let exported = export( + &document, + &WebExportOptions::default(), + Workflow::HtmlAndImages, + ) + .unwrap(); + let html = exported.html.as_deref().unwrap(); + + assert!(html.contains("onMouseOver=\"MM_swapImage('left','','right.png',1);\"")); + assert!(html.contains("onload=\"MM_preloadImages('right.png')\"")); + assert!( + exported.files.iter().any(|file| file.name == "right.png"), + "the swap target is among the exported files" + ); +} + #[test] fn expv_overrides_choose_the_codec() { let mut document = sliced_document(); diff --git a/docs/HANDOFF.md b/docs/HANDOFF.md index f8443001..58d9f02c 100644 --- a/docs/HANDOFF.md +++ b/docs/HANDOFF.md @@ -402,17 +402,26 @@ the absolute repo path. `E0425: cannot find function 'menu_census' in module 'app::menus'` while the run's own commit carries **both** the definition (`crates/app/src/menus.rs`) and the call (`xtask/src/climb.rs`) — - so the tree was consistent and the *artifact* was not. The - mechanism is not yet identified (the pipeline declares no - `cache:` mount, the CI image bakes no target directory, and the - runner creates a fresh container for every workspace job), so the - owning component is **GitForge CI**, not this repository. + so the tree was consistent and the *artifact* was not. + **The mechanism is identified** (decision + [0122](decisions/0122-the-cargo-cache-is-scoped-per-job-not-per-commit.md)): + the pipeline's per-job `cache:` mount is scoped to the *job*, not + to the commit, so cargo can reuse another branch's `app` rlib. The + mitigation is in the tree — `scripts/ci_refresh_stale_crates.sh` + runs before any job compiles and refreshes only the crates whose + sources are newer than the cached copy. The last recurrence was + the `layers` job of run `5c64a539` (2026-11-16T15:05:32Z); the + mitigation landed in `702ad246` (2026-11-17T02:29Z) and the + store records **44 runs and 504 jobs with no recurrence since**. + The platform defect itself (the cache's scope) stays open with + the GitForge platform lane as its owner. **What to do:** do not read the red as a source defect and do not "fix" the call site; confirm the commit is self-consistent (`git grep '' -- crates/app`), then re-trigger the run (an empty commit re-checks-out the tree and the job goes - green). Owner and the open question are recorded on the board - (`docs/TASKS.md`, CI/tooling lane). + green). If the mitigation's step is *absent* from the job's log, + the run predates the mitigation — that is a different situation + from a recurrence. - **Local gates before any push:** the full gate is `cargo run -p xtask -- ci` (fmt, clippy `-D warnings`, tests, layers, wasm, corpus, goldens, perf, fuzz-quick, parity, scorecard, diff --git a/docs/TASKS.md b/docs/TASKS.md index 1d8ea089..0c4a2b47 100644 --- a/docs/TASKS.md +++ b/docs/TASKS.md @@ -77,10 +77,10 @@ Review follow-up: [filter controls and master transform](tasks/filter-controls-r |---|---|---|---|---|---| | 5 | 19.6 inert-menu wiring campaign | [tasks/T5-inert-burndown.md](tasks/T5-inert-burndown.md) | `DONE — 2026-11-11` | UI/verb lane | Batches 1–17 are complete. The current pinned census is 2 inert rows of 298; eight other rows remain explicitly feature/evidence-gated and are not part of an active wiring campaign (`menus::tests::inert_count_never_rises`, `menus::tests::menu_item_total_is_pinned`). | | 194 | Aegis gate: content-keyed baseline, so a doc edit no longer fails the gate (decision 0074) | [decisions/0074-aegis-content-keyed-baseline.md](decisions/0074-aegis-content-keyed-baseline.md) | `IMPLEMENTED` — the content-keyed gate and its decision record are in the tree; the gate has been exercised on this branch's tree | CI/tooling lane | **landed in `main` (#18)** and merged here; attach a run id before claiming the fleet-wide state | -| — | CI false red: a job compiles against an `app` artifact built from another tree (`E0425: cannot find function 'menu_census' in module 'app::menus'`) | [HANDOFF §6.1](HANDOFF.md#61-pipeline) | `OPEN — DEFECT, NOT A SOURCE FAULT` — measured 2026-11-16 on six branch runs (`a329bda0`, `e2e85b9b`, `57660fb8`, `5e48f383`, `b28fd3f5`, `0757082e`, `a6b23df6`); each run's own commit carries both the definition and the call, so the tree was consistent and the artifact was not. The mechanism is not yet identified: this pipeline declares no `cache:` mount, the CI image bakes no target directory, and the runner creates a fresh container per workspace job | **GitForge CI** (not this repository) | identify where the stale `app` rlib enters the job — the runner's artifact handling is the lead — and fix it there; until then, a run that fails with this signature is re-triggered (an empty commit) rather than read as a source defect | -| — | Release workflow hardening: every action pinned to a commit SHA, a fail-closed pin check in CI (decision 0099), a classified release mode with a fail-closed aggregate gate (decision 0100), a pinned measuring toolchain (decision 0102) | [decisions/0099-workflow-actions-are-pinned.md](decisions/0099-workflow-actions-are-pinned.md) · [decisions/0100-release-mode-classification-and-a-tested-release-gate.md](decisions/0100-release-mode-classification-and-a-tested-release-gate.md) · [decisions/0102-coverage-floor-is-measured-with-a-pinned-toolchain.md](decisions/0102-coverage-floor-is-measured-with-a-pinned-toolchain.md) | `IMPLEMENTED` — the 19 action references are pinned; the aggregate gate is a tested script that refuses a missing, empty, extra, mismatched or platform-less artifact set; a rehearsal is a `TEST`-titled prerelease on a disposable tag; the validator and the three release self-tests pass locally; the CI receipt is pending | CI/tooling lane | land the branch on both receipts; the checks run in the backup CI's `changes` job, and mirroring them into the GitForge pipeline waits for the `:2` pipeline shape (PR #123, PR #129); then rehearse the release path end to end through `workflow_dispatch` | +| — | CI false red: a job compiles against an `app` artifact built from another tree (`E0425: cannot find function 'menu_census' in module 'app::menus'`) | [HANDOFF §6.1](HANDOFF.md#61-pipeline) | `RESOLVED — the mechanism is identified and the mitigation is in the tree` — the mechanism is the cache's **scope** (decision [0122](decisions/0122-the-cargo-cache-is-scoped-per-job-not-per-commit.md)): the pipeline's per-job `cache:` mount is not scoped to the commit, so cargo can reuse another branch's `app` rlib. `scripts/ci_refresh_stale_crates.sh` now runs before any job compiles, and refreshes only the crates whose sources are newer than the cached copy; its own self-test lays the mtimes out and proves the selection. The last recurrence of this signature was the `layers` job of run `5c64a539` (2026-11-16T15:05:32Z); the mitigation landed in `702ad246` (2026-11-17T02:29Z). **Since then the store records 44 runs and 504 jobs with no recurrence.** The one later `E0425` is a *different* signature — `cannot find function 'stated_tag_names_version' in this scope` in the xtask manifest test (the `clippy-test` job of `c396b18a`, 2026-11-17T22:08:03Z), a genuine source fault, fixed by `d1029347` ("the manifest tests import the helpers they exercise"); it is not this class | CI/tooling lane — the root cause is the runner's cache scope (GitForge CI); the mitigation lives in this repository (`.gitforge.yml` + `scripts/ci_refresh_stale_crates.sh`) | closed — the mechanism is identified, the mitigation is in the tree and measured, and no recurrence followed; a future run that fails with this signature is a new defect, not this one | +| — | Release workflow hardening: every action pinned to a commit SHA, a fail-closed pin check in CI (decision 0099), a classified release mode with a fail-closed aggregate gate (decision 0100), a pinned measuring toolchain (decision 0102) | [decisions/0099-workflow-actions-are-pinned.md](decisions/0099-workflow-actions-are-pinned.md) · [decisions/0100-release-mode-classification-and-a-tested-release-gate.md](decisions/0100-release-mode-classification-and-a-tested-release-gate.md) · [decisions/0102-coverage-floor-is-measured-with-a-pinned-toolchain.md](decisions/0102-coverage-floor-is-measured-with-a-pinned-toolchain.md) | `IMPLEMENTED` and **rehearsed end to end** — the 19 action references are pinned; the aggregate gate is a tested script that refuses a missing, empty, extra, mismatched or platform-less artifact set; a rehearsal is a `TEST`-titled prerelease on a disposable tag; the validator and the three release self-tests pass locally. The CI receipt exists: `workflow_dispatch` runs **44188107435** (`v0.2.5-dryrun.6`, head `c396b18a`) completed `success`, produced draft 449279892 with nine uploaded assets, and passed the aggregate gate and the mode classifier — [receipt](release/RECEIPT-0.2.5-DRYRUN-6.md) | CI/tooling lane | publish the production release from a real tag, which needs explicit maintainer authorization (campaign §20); until then, the rehearsal drafts and their disposable tags stay on the remote and their removal is maintainer-gated | | — | The coverage gate explains itself: the per-file breakdown, each file's `DA` record count beside its `LF`, and the profile inventory (decision 0103) | [decisions/0103-coverage-export-counts-a-file-once-per-build.md](decisions/0103-coverage-export-counts-a-file-once-per-build.md) | `IMPLEMENTED` on the branch; the CI receipt is pending | CI/tooling lane | land it on both receipts; the check that closes the decision is the next pinned run, where `LF` must equal the `DA` count again for `crates/app/src/shell/mod.rs` — under rustc 1.100.0 the export reported 8138 lines for a file that has 6309 | -| 27 | Cross-platform release engineering: Windows, macOS, Linux package matrix, Android feasibility, source, SBOM/provenance, signing, CI aggregation and certification | [tasks/T27-cross-platform-release.md](tasks/T27-cross-platform-release.md) · the atomic ledger is [release/EXECUTION-BOARD.md](release/EXECUTION-BOARD.md) (62 rows, one per `docs/tasks/T27-*.md`) | `IN PROGRESS` — implementation is ahead of verification; **no target is `CERTIFIED`** and nothing is signed; the SBOM half is build-verified against the vendored CycloneDX 1.5 schema and compared against the lockfile's own inventory (decision [0119](decisions/0119-the-sbom-is-validated-against-the-schema-it-claims.md)) | release-workflow lane | land an end-to-end tag run of the repaired workflow and attach the run id as the RP-07 receipt; consolidate the duplicated `dist --source`/`--sbom` verbs (B-3); the board ledger was reconciled on 2026-11-16 (T27-59) and again on 2026-11-17 — 62 rows against 62 tickets, no problems; PR #8's undispositioned comments are adjudicated (T27-24) and the one valid application defect among them — a plotted draft carrying no owning tool — is fixed on `main` by `d3e45cf9` (T27-23, PR #171), and the Aegis scan now runs on documentation-only changes in both CI systems (T27-62, PR #172 → `7bcf28fd`) | +| 27 | Cross-platform release engineering: Windows, macOS, Linux package matrix, Android feasibility, source, SBOM/provenance, signing, CI aggregation and certification | [tasks/T27-cross-platform-release.md](tasks/T27-cross-platform-release.md) · the atomic ledger is [release/EXECUTION-BOARD.md](release/EXECUTION-BOARD.md) (62 rows, one per `docs/tasks/T27-*.md`) | `IN PROGRESS` — implementation is ahead of verification; **no target is `CERTIFIED`** and nothing is signed; the SBOM half is build-verified against the vendored CycloneDX 1.5 schema and compared against the lockfile's own inventory (decision [0119](decisions/0119-the-sbom-is-validated-against-the-schema-it-claims.md)); the **RP-07 receipt now exists** — the end-to-end tag run of the repaired workflow is run **44188107435** (`v0.2.5-dryrun.6`), which produced draft 449279892 with nine assets, ran the web bridge smoke green, passed the DEB's eight-stage package verification in a clean Debian 12 container, and had three of its assets independently re-verified from the draft ([receipt](release/RECEIPT-0.2.5-DRYRUN-6.md)); the scope that can honestly be certified is written up in [release/CERTIFICATION.md](release/CERTIFICATION.md) | release-workflow lane | publish the production release from a real tag — maintainer-authorized only (campaign §20); consolidate the duplicated `dist --source`/`--sbom` verbs (B-3); the board ledger was reconciled on 2026-11-16 (T27-59) and again on 2026-11-17 — 62 rows against 62 tickets, no problems, with eleven rows advanced by the rehearsals (T27-19/21/49/50/51/52/53/54/55/57/60) and a twelfth (T27-25) advanced by the green CI on the head `f710c524` — the full gate, the application test suite, the web build and the browser smoke all pass there; PR #8's undispositioned comments are adjudicated (T27-24) and the one valid application defect among them — a plotted draft carrying no owning tool — is fixed on `main` by `d3e45cf9` (T27-23, PR #171), and the Aegis scan now runs on documentation-only changes in both CI systems (T27-62, PR #172 → `7bcf28fd`) | | 26 | Refresh user-visible CS6 parity matrix | [tasks/T26-parity-matrix.md](tasks/T26-parity-matrix.md) | `IMPLEMENTED` — matrix published | parity lane | every CS6 menu, tool, panel, inspector field set, workflow, import/export surface, canvas interaction, keymap area, and selection state is classified with an evidence citation and an owning ticket in [`CS6-PARITY.md`](CS6-PARITY.md); the browser-sampling check the ticket named was performed on 2026-11-04 against a fresh optimized bundle and is recorded in T22 | | 22 | Phase 22: CS6 shell fidelity and screenshot audit | [tasks/T22-screenshot-ui.md](tasks/T22-screenshot-ui.md) | `IN PROGRESS` | UI lane | screenshot acceptance is complete; 22.1 panel inventory and 22.2 tool strip landed on `main` with the shell-lineage train (#107); 22.5 chrome remains partial, and 22.4 still has evidence-gated bitmap transform/fill-stroke work. [Phase 22](NEXT.md#phase-22--cs6-shell-fidelity-p0-the-looks-and-acts-like-fireworks-phase) | | 20 | 20.1 In-canvas text editing (egui TextEdit overlay; kittest focus tests) | [tasks/T20-text-editing.md](tasks/T20-text-editing.md) | `IMPLEMENTED` — landed 2026-11-06 (`shell::begin_text_edit` / `text_edit_overlay`) | editing lane | Phase 20's other items are tracked in the rows above and in the Phase 20 board below | @@ -97,6 +97,7 @@ Review follow-up: [filter controls and master transform](tasks/filter-controls-r | # | Task | Ticket | |---|---|---| | — | 20.4 Mask stack UI (add/remove/mode/link/move) — `PARTIAL`: the stack, its fold, and the level panel landed in #113 (decisions 0092/0093) with the level add/remove/move affordances; the screenshot acceptance and the compositor's mask sampling resolution remain ([#114](https://github.com/aliasfoxkde/FireCraft/issues/114)) — a FireCraft extension, not a CS6 parity claim | [tasks/T9-mask-stack.md](tasks/T9-mask-stack.md) | +| — | 20.7 Slice behaviors + HTML preview — **the export-level receipt landed** (`crates/app-cli/tests/web_behavior_export.rs`): a behavior attached through the engine's command surface reaches the exported HTML as the `MM_*` markup, and the file it points at is one the export wrote. Writing the regression exposed a real defect — the model stores the *name* of the slice a swap points at, while the exporter emitted that name where the runtime needs a *file* — fixed by resolving the name to the exported file at export time ([decision 0133](decisions/0133-a-behavior-target-names-a-slice-and-the-exporter-resolves-the-file.md)). The preview's own widgets are now asserted through egui's own harness (`crates/app/src/shell/html_preview.rs`, `the_preview_draws_its_widgets_through_the_harness`): the panel is drawn into the accessibility tree and its labels are queried as widgets. The image-snapshot half of that coverage needs egui_kittest's `wgpu` feature and a verified Vulkan adapter, which this host and the CI image do not carry — that half is the queued item 21 below ([#93](https://github.com/aliasfoxkde/FireCraft/issues/93) stays open on it) | — | 20.1 and 20.5–20.7 are already on `main`; do not reopen them. diff --git a/docs/decisions/0122-the-cargo-cache-is-scoped-per-job-not-per-commit.md b/docs/decisions/0122-the-cargo-cache-is-scoped-per-job-not-per-commit.md index d7db7505..d1a83265 100644 --- a/docs/decisions/0122-the-cargo-cache-is-scoped-per-job-not-per-commit.md +++ b/docs/decisions/0122-the-cargo-cache-is-scoped-per-job-not-per-commit.md @@ -142,7 +142,16 @@ crates and refreshed nothing. The reference must be the fingerprint - A branch's CI verdict no longer depends on which other branch ran before it. The campaign branch is re-run on top of this step; its - `layers` job is the receipt. + `layers` job is the receipt. **Measured 2026-11-17**: the last + recurrence of this failure class was the `layers` job of run + `5c64a539` (2026-11-16T15:05:32Z), and this mitigation landed in + `702ad246` (2026-11-17T02:29Z). The GitForge job store records + **44 runs and 504 jobs** after that landing with **no recurrence**. + (A later `E0425` — `cannot find function 'stated_tag_names_version' + in this scope`, the `clippy-test` job of `c396b18a` — is a different + signature: a genuine source fault in the xtask manifest test, fixed + by `d1029347`. It is not this class and is not evidence against + this mitigation.) - The cost is bounded and paid only when the cache is actually contended: in the contended case the local crates are rebuilt, which is the price of a verdict that means something. Against the real diff --git a/docs/decisions/0133-a-behavior-target-names-a-slice-and-the-exporter-resolves-the-file.md b/docs/decisions/0133-a-behavior-target-names-a-slice-and-the-exporter-resolves-the-file.md new file mode 100644 index 00000000..301fa30b --- /dev/null +++ b/docs/decisions/0133-a-behavior-target-names-a-slice-and-the-exporter-resolves-the-file.md @@ -0,0 +1,106 @@ +# 0133 — A behavior's target names a slice; the exporter resolves the file it wrote + +Date: 2026-11-17 · Tickets: Phase 20.7 (issue #93) · Status: accepted + +## Context + +A slice behavior carries its parameters in the slice's `DTAv` custom-data +map under the key `swapTarget` (decision 0017). Two surfaces write that +key, and they write the *same* kind of value: + +- The engine's `slice.addSwapImageBehaviorFromPoint` — CS6's "add swap + image behavior from point" verb — stores the **name** of the slice + under the point. +- The shell's Behaviors panel writes the same string: its target + dropdown lists the document's other slices **by name** + (`panels_web::swap_targets`). + +The exporter, meanwhile, emitted that value verbatim into the markup: + +```html + + +``` + +The Dreamweaver runtime those calls imitate sets `img.src` from the third +argument and `new Image().src` from the preload list, so both must name +a **file the export wrote**. The export writes `right.png`. + +The names the export writes are export-time artifacts: decision 0017 +records that the auto-naming scheme is applied at export time and is +**never stored on the slice**. So the model cannot hold the file name +either — the file name is not known until the export lays the grid out, +names the pieces and encodes them. + +The result was that every behavior attached through the command surface +— the only way the shell attaches one — produced HTML that points at a +file that does not exist. The swap silently did nothing in a browser and +the preload 404'd. The defect was invisible to the tests: the exporter's +own behavior test (`webexport`'s `behaviors_emit_hooks_and_the_runtime`) +built the behavior **directly on the model** with a hand-written file +name (`swap.png`), a value the command surface never writes, so it +asserted a case the product does not produce. + +## Decision + +1. **The model stores the name of the slice a behavior points at.** The + engine's verbs and the shell's dropdown keep writing names; the file + a swap points to stays an export-time artifact. +2. **The exporter resolves that name to the file it wrote.** The + resolution map is keyed by every name the object can be addressed by: + the slice's object name, its internal name, and each of those + cleaned. The value is the object's first piece's file — the file the + export actually produced for it. +3. **A target that names no slice is a file name the user typed** and is + emitted verbatim, so a hand-authored `swap.png` still works and the + exporter's existing markup is unchanged for it. +4. **A file the export wrote is referenced where it was written.** The + resolved target and the preload list carry the export's + `images_dir` prefix, exactly as the `` that names the same + file does. A file the user typed is their own reference and is used as + given. + +## Why + +The alternative — storing the file name on the slice — cannot be built +without breaking two rules at once. The name depends on the export's +grid layout and on the slice's `EXPv` codec, so the engine would have to +reach the exporter's naming rules to compute it; `engine` is L3 and +`webexport` is L4, and an upward edge is a layer violation. And storing +the name would freeze an export-time artifact on the slice, which is the +rule decision 0017 set. + +Resolving at the exporter keeps the model honest (it names a slice, and +a slice can be renamed — the swap follows) and keeps the exporter's +promise (the markup references files that exist, at the paths they were +written). + +## Consequences + +- The shell's own workflow now produces working swaps: pick a target + slice in the dropdown, and the exported page swaps to that slice's + image. +- The exporter's markup for a hand-written file name target is + unchanged (`swap.png` still emits `swap.png`), so no existing golden + moves. +- The export-level regression that closes issue #93 drives the behavior + through the command surface (`slice.create`, `selection.set`, + `slice.addSwapImageBehaviorFromPoint`, `slice.addBehavior`, + `slice.setData`) and asserts the emitted HTML carries the behavior's + markup — and that the file it points at is among the exported files. +- A target naming a slice that the export did not write (a slice off the + grid, or one the layout dropped) falls back to the verbatim name: the + markup then names the slice, not a file. That is the same shape as a + hand-written target, and it is visible in the exported HTML rather + than silent. + +## References + +- `crates/webexport/src/html.rs` — `target_files` (the resolution map) + and `hooks` (the markup). +- `crates/engine/src/catalog/web.rs` — + `slice.addSwapImageBehaviorFromPoint`, which writes the name. +- `crates/app-cli/tests/web_behavior_export.rs` — the export-level + regression, driven through the command surface. +- [decision 0017](0017-web-layer-and-export.md) — the Web Layer, the + `DTAv` parameter map and export-time naming. diff --git a/docs/release/ARTIFACT-MATRIX.md b/docs/release/ARTIFACT-MATRIX.md index fccbd418..ebaa0502 100644 --- a/docs/release/ARTIFACT-MATRIX.md +++ b/docs/release/ARTIFACT-MATRIX.md @@ -1,6 +1,6 @@ # Release artifact matrix — RP-00 inventory (T27) -> **Reconciled 2026-11-09** against the Phase 0 baseline audit +> **Reconciled 2026-11-17** against the Phase 0 baseline audit > ([PHASE-0-AUDIT.md](PHASE-0-AUDIT.md), PR #16, branch > `docs/phase-0-baseline-audit`). The audit is the evidence authority > for what exists; [decision 0075](../decisions/0075-release-scope-mvp.md) @@ -42,22 +42,22 @@ | # | Target | Format | What exists today | How it is produced | Status | Rung | Receipt | Blocker | |---|---|---|---|---|---|---|---|---| -| 1 | Linux x86_64 | tar.gz (desktop) | `firecraft-0.2.5-linux-x86_64.tar.gz` | `xtask dist --native` → `dist/native/` | `PUBLISHED` | 11 | GitHub release v0.2.5 asset, 22,084,726 B. Downloaded and checksum-verified 2026-11-09; `sha256sum -c` passes against the published `MANIFEST.sha256`; unpacks to `firecraft-0.2.5/bin/{firecraft,firecraft-cli}` + `share/licenses/LICENSE` | B-18 (a rebuild of the current tree does not reproduce these bytes), B-21 (published out of band, not by the release workflow) | -| 2 | Linux x86_64 | tar.gz (CLI) | `firecraft-cli-0.2.5-linux-x86_64.tar.gz` | `xtask dist --native` → `dist/native/` | `PUBLISHED` | 11 | GitHub release v0.2.5 asset, 8,072,377 B. Downloaded and checksum-verified 2026-11-09; the extracted `firecraft-cli` **launches** and prints its usage | B-18, B-21 | -| 3 | Linux x86_64 | DEB | `firecraft_0.2.5_amd64.deb` | `xtask dist --native` from the pinned `packaging/deb/control`; the gate reads the package back with `dpkg-deb` and fail-closes on the declared libc6 floor against the binaries' max GLIBC symbol version (via `objdump`) and on the presence of the desktop entry and icon in the package contents | `PUBLISHED` | 11 | GitHub release v0.2.5 asset, 14,819,904 B. Downloaded and checksum-verified 2026-11-09; `dpkg-deb --info` reads `Version: 0.2.5`, `Architecture: amd64`; the extracted `firecraft-cli` launches. **Clean-container receipts on the published v0.2.3 package** (Debian 12): install → `PASS: installed firecraft 0.2.3`; CLI document round trip → `firecraft-cli info` prints the document JSON, `firecraft-cli export` writes a 97 B PNG with the correct signature `89504e470d0a1a0a`; GUI launch under Xvfb → exit 124, i.e. the binary started and stayed up for the 25 s launch window | **B-7 — OPEN DEFECT.** The declared glibc floor is below the binaries' real requirement, confirmed on the **published** artifact by `objdump -p`, not only on a local rebuild. The floor numbers and the `objdump` floor logic are owned by the stacked `fix/deb-glibc-floor` PR; this row deliberately does not restate them. A package built from this tree under-declares its dependency, so it installs cleanly and then fails to start on older distributions. **This is a defect, not a settled fact.**
**B-24 — OPEN DEFECT (owner: PR #41).** The published package carries **no desktop integration at all**: `dpkg-deb --contents` on the published v0.2.5 lists only `./usr/bin/firecraft`, `./usr/bin/firecraft-cli` and `./usr/share/doc/firecraft/copyright`. The install verification run against the published v0.2.3 package in a clean Debian 12 container fails at exactly that point — `FAIL: /usr/share/applications/firecraft.desktop is not installed` — so the package installs cleanly and is still not a desktop application. PR #41 adds the desktop entry and the icon to the payload and makes `dist --native` fail when either is absent from the built package The published package's missing desktop entry and icon are also tracked as board T27-22. | +| 1 | Linux x86_64 | tar.gz (desktop) | `firecraft-0.2.5-linux-x86_64.tar.gz` | `xtask dist --native` → `dist/native/` | `PUBLISHED` | 11 | GitHub release v0.2.5 asset, 22,084,726 B. Downloaded and checksum-verified 2026-11-09; `sha256sum -c` passes against the published `MANIFEST.sha256`; unpacks to `firecraft-0.2.5/bin/{firecraft,firecraft-cli}` + `share/licenses/LICENSE`. **The workflow's own build path now produces this artifact**: the `v0.2.5-dryrun.6` rehearsal published `firecraft-0.2.5-linux-x86_64.tar.gz` (24,389,908 B, sha256 `cb55a525…`) on **draft** 449279892 — a draft is not a publication, so those bytes are evidence of a working build, not of a published asset; they are also **byte-identical across two independent rehearsals** (dryrun.5 and dryrun.6) | B-18 (a rebuild of the current tree does not reproduce the *published* bytes — the rehearsal's tarball is 2,305,182 B larger), B-21 (published out of band, not by the release workflow) | +| 2 | Linux x86_64 | tar.gz (CLI) | `firecraft-cli-0.2.5-linux-x86_64.tar.gz` | `xtask dist --native` → `dist/native/` | `PUBLISHED` | 11 | GitHub release v0.2.5 asset, 8,072,377 B. Downloaded and checksum-verified 2026-11-09; the extracted `firecraft-cli` **launches** and prints its usage. **The workflow's own build path now produces this artifact, and it attests its own build**: the `v0.2.5-dryrun.6` rehearsal published `firecraft-cli-0.2.5-linux-x86_64.tar.gz` (8,158,124 B, sha256 `953633e4…`) on **draft** 449279892; the archive was downloaded from that draft, re-hashed locally (the digest matches) and extracted, and the extracted binary prints `firecraft-cli 0.2.5`. It is byte-identical across two independent rehearsals | B-18, B-21 (the production release's bytes were published out of band) | +| 3 | Linux x86_64 | DEB | `firecraft_0.2.5_amd64.deb` | `xtask dist --native` from the pinned `packaging/deb/control`; the gate reads the package back with `dpkg-deb` and fail-closes on the declared libc6 floor against the binaries' max GLIBC symbol version (via `objdump`) and on the presence of the desktop entry and icon in the package contents | `PUBLISHED` | 11 | GitHub release v0.2.5 asset, 14,819,904 B. Downloaded and checksum-verified 2026-11-09; `dpkg-deb --info` reads `Version: 0.2.5`, `Architecture: amd64`; the extracted `firecraft-cli` launches. **Clean-container receipts on the published v0.2.3 package** (Debian 12): install → `PASS: installed firecraft 0.2.3`; CLI document round trip → `firecraft-cli info` prints the document JSON, `firecraft-cli export` writes a 97 B PNG with the correct signature `89504e470d0a1a0a`; GUI launch under Xvfb → exit 124, i.e. the binary started and stayed up for the 25 s launch window | **B-7 — OPEN DEFECT.** The declared glibc floor is below the binaries' real requirement, confirmed on the **published** artifact by `objdump -p`, not only on a local rebuild. The floor numbers and the `objdump` floor logic are owned by the stacked `fix/deb-glibc-floor` PR; this row deliberately does not restate them. A package built from this tree under-declares its dependency, so it installs cleanly and then fails to start on older distributions. **This is a defect, not a settled fact.**
**B-24 — OPEN DEFECT (owner: PR #41).** The published package carries **no desktop integration at all**: `dpkg-deb --contents` on the published v0.2.5 lists only `./usr/bin/firecraft`, `./usr/bin/firecraft-cli` and `./usr/share/doc/firecraft/copyright`. The install verification run against the published v0.2.3 package in a clean Debian 12 container fails at exactly that point — `FAIL: /usr/share/applications/firecraft.desktop is not installed` — so the package installs cleanly and is still not a desktop application. PR #41 adds the desktop entry and the icon to the payload and makes `dist --native` fail when either is absent from the built package The published package's missing desktop entry and icon are also tracked as board T27-22. **The fix is in the tree and the workflow's own build proves it**: the `v0.2.5-dryrun.6` rehearsal published `firecraft_0.2.5_amd64.deb` (16,228,104 B, sha256 `79283696…`) on **draft** 449279892, and the release workflow's package verification ran the repository's eight-stage script against that package in a clean Debian 12 container — **all eight stages passed**, including the desktop entry, the icon and the **upgrade** stage that the distro matrix does not exercise. The gap is therefore between the *published* package and the tree, not inside the tree | B-7 and B-24 remain defects **of the published v0.2.5 package**, which this tree no longer reproduces; they close only when a release is cut from this tree | | 4 | Linux x86_64 | AppImage | recipe only — `packaging/build-appimage.sh` | operator-run on a host with `linuxdeploy`; the script is `bash -n` parse-checked by `dist --native` | `NOT STARTED` | 1 | recipe exists and passes the gate's `bash -n` check; **never executed**. On this branch the recipe gained a preflight that fails when the desktop entry or the icon is missing from the tree, so the "no icon → an image with no icon" silent defect cannot reach `linuxdeploy` (audit §11.6) | B-8 (`linuxdeploy` absent, no pinned copy vendored), B-10 (the script stages `target/release/…` and ignores `CARGO_TARGET_DIR`) B-9 (the icon file and the `Icon=` key are in-tree since `a4d74a0`, but `linuxdeploy` has still never run). | | 5 | Linux x86_64 | Flatpak | recipe only — `packaging/build-flatpak.sh` + `packaging/flatpak/dev.firecraft.FireCraft.yml` (both added on this branch; **no recipe on `main`**) | operator-run on a host with the Flatpak SDK; the script is `bash -n` parse-checked and the manifest is YAML-parse-checked by `dist --native` | `NOT STARTED` | 1 | recipe and manifest exist on this branch and the gate parse-checks them; **never built**. On this branch the recipe no longer passes the unsupported `flatpak run --repo` argument, installs the bundle into a user installation, launches it with `flatpak run --command=firecraft-cli`, removes it and fails if it is still listed; the manifest installs the icon and points at the canonical desktop entry (audit §11.6) | B-11 (`flatpak` and `flatpak-builder` absent) The manifest's module source is `type: dir, path: ..` — the local working tree, not an immutable upstream revision — so a bundle built from it is not reproducible from the tag SHA (board T27-29). | | 6 | Windows x86_64 | portable ZIP | recipe only — `packaging/build-windows.ps1` | operator-run on a Windows host with MSVC | `NOT STARTED` | 1 | recipe exists; the gate **existence-checks** it only — it cannot parse PowerShell on this fleet; **never executed** | B-12 (`pwsh` absent; no MSVC linker on this host). A GNU cross-build and a wine smoke are recorded on the PR #9 branch — preview evidence only, never native certification (campaign §8) The required Windows artifact does not yet exist: the MSVC build, the installer and the clean-Windows-VM install are all outstanding (board T27-36/37/38). | | 7 | Windows x86_64 | Inno Setup installer | recipe only — `packaging/windows/firecraft.iss` | compiled by the PowerShell recipe on a Windows host with Inno Setup 6 | `NOT STARTED` | 1 | recipe exists and is existence-checked by the gate; the `__APPVERSION__` placeholder is substituted by the PowerShell recipe, which has never run; **never built** | B-12 (`ISCC` absent). The PowerShell recipe *skips* the installer when `ISCC` is missing, so a Windows run without Inno Setup ships no installer and says nothing | | 8 | macOS universal | DMG | recipe only — `packaging/build-macos-dmg.sh` | operator-run on macOS: two `apple-darwin` slices → `lipo` → `hdiutil` | `NOT STARTED` | 1 | recipe exists and is `bash -n` parse-checked by the gate; **never executed** | needs macOS + Xcode CLT; `lipo`/`hdiutil` are absent on this Linux host, so the recipe is off-fleet by design | -| 9 | Web (WASM) | ZIP bundle | `firecraft-web-0.2.5.zip` | `xtask dist --web` → `dist/web/`, then zipped | `PUBLISHED` | 11 | GitHub release v0.2.5 asset, 7,174,826 B. Downloaded 2026-11-09: 6 entries (bundle `firecraft-web-f9dbe2f14305b2d9{,_bg.wasm}`, `index.html`, `host.html`, `appinfo.json`, inner `MANIFEST.sha256`); `sha256sum -c` passes on all 5 manifest entries | **B-21 / F-16 — the published bundle was never built by the release workflow.** Its entries are stamped 2026-11-03 12:08, six hours before the workflow's own build finished at 18:15:59Z, and its wasm is a different size than the workflow's. No recorded bridge smoke ever ran against the published bytes. Separately, the **live deployment** serves a fourth bundle (`2274ac8c2925b61d`) that matches no release — B-6; the live comparison was **re-verified 2026-11-10** and is recorded in [Live web deployment](#live-web-deployment-b-6-reconciliation) below | +| 9 | Web (WASM) | ZIP bundle | `firecraft-web-0.2.5.zip` | `xtask dist --web` → `dist/web/`, then zipped | `PUBLISHED` | 11 | GitHub release v0.2.5 asset, 7,174,826 B. Downloaded 2026-11-09: 6 entries (bundle `firecraft-web-f9dbe2f14305b2d9{,_bg.wasm}`, `index.html`, `host.html`, `appinfo.json`, inner `MANIFEST.sha256`); `sha256sum -c` passes on all 5 manifest entries | **B-21 / F-16 — the published bundle was never built by the release workflow.** Its entries are stamped 2026-11-03 12:08, six hours before the workflow's own build finished at 18:15:59Z, and its wasm is a different size than the workflow's. No recorded bridge smoke ever ran against the published bytes. **The workflow's own build path now produces this artifact, and the smoke ran against it**: the `v0.2.5-dryrun.6` rehearsal published `firecraft-web-0.2.5.zip` (5,853,417 B, sha256 `898be851…`) on **draft** 449279892, and the `build artifacts` job's log reads `[xtask] dist web: 7 file(s), wasm 15780814 B …, smoke green` and `[xtask] dist --web ok`. The ZIP's **payload is byte-identical** across two independent rehearsals (all seven files hash the same); only the ZIP container's own timestamps differ. Separately, the **live deployment** serves a fourth bundle (`2274ac8c2925b61d`) that matches no release — B-6; the live comparison was **re-verified 2026-11-10** and is recorded in [Live web deployment](#live-web-deployment-b-6-reconciliation) below | | 10 | WebOS | `appinfo.json` + `host.html` | staged into `dist/web/` by `dist --web` from `packaging/webos/` | `xtask dist --web` | `BUILD VERIFIED` | 5 | staged `appinfo.json` reads 0.2.5; the web manifest covers `appinfo.json` and `host.html`; the live site serves the same `appinfo.json` bytes — re-verified 2026-11-10, byte-identical (`7d039c36…`) | device/runtime smoke is `UNVERIFIED` (audit §6); the Pages deploy is manual and currently stale (B-6) | | 11 | CLI binaries | `firecraft-cli` (inside #1 and #2) | `firecraft-cli` | `xtask dist --native` | `PUBLISHED` | 11 | the **published** `firecraft-cli` was downloaded and extracted 2026-11-09 and **launches**, printing its full usage string; the locally rebuilt one does the same | — | -| 12 | Source | tar.gz | `firecraft-0.2.5-source.tar.gz` — the canonical name; **not on `main`** | `xtask dist --source` (git archive of HEAD; the verb verifies a clean extraction and runs `cargo metadata --locked` inside it). The verb exists on this branch and on the PR #9 / #12 trees | `BUILD VERIFIED` | 3 | produced 2026-11-09 on `docs/release-execution-board` @ `08eae4a`: 3.35 MB, 661 entries, sha256 `a0fcaafe…`; clean-extract + `cargo metadata --locked` verified inside the verb. **That build predates the rename**: it shipped as `FireCraft-0.2.5-source.tar.gz`, and the rehearsal (v0.2.5-dryrun.3) exposed the capitalised name as a naming defect. `xtask/src/dist.rs` now emits the lowercase name; the row's receipt is not yet re-earned against it (see [RECEIPT-0.2.5-DRYRUN-3.md](RECEIPT-0.2.5-DRYRUN-3.md), T27-52) | F-9 (the verb is absent from `main`), B-13 (never published) | -| 13 | SBOM | CycloneDX 1.5 JSON | `firecraft-0.2.5-sbom.cdx.json` — **not on `main`** | `xtask dist --sbom` over the locked dependency graph; the serial is a deterministic UUIDv8 derived from the locked graph. The verb exists on this branch and on the PR #9 / #12 trees | `BUILD VERIFIED` | 3 | produced 2026-11-09 on `docs/release-execution-board` @ `08eae4a`: 249 KB, 493 components, 494 dependency edges, sha256 `837a731b…` | F-9 (the verb is absent from `main`), B-13 (never published) | -| 14 | Metadata | checksums / manifest | `dist/native/MANIFEST.sha256` (3 native artifacts) and `dist/web/MANIFEST.sha256` (5 bundle files) | `xtask dist --native` / `dist --web`; `release.yml` additionally writes a release-wide `RELEASE_SHA256.txt` over 8 assets | `PUBLISHED` | 11 | `MANIFEST.sha256` attached to release v0.2.5 (398 B, 3 native files) and embedded in the web ZIP (532 B). Both checked against the published downloads on 2026-11-09 — 3/3 and 5/5 pass | B-17 (neither list covers the ZIP files themselves and no release-wide `RELEASE_SHA256.txt` was ever published, although `release.yml` writes one), B-21 (the manifest and the assets it covers were published out of band, so the list is self-consistent but not independently attested) | +| 12 | Source | tar.gz | `firecraft-0.2.5-source.tar.gz` — the canonical name; **not on `main`** | `xtask dist --source` (git archive of HEAD; the verb verifies a clean extraction and runs `cargo metadata --locked` inside it). The verb exists on this branch and on the PR #9 / #12 trees | `SMOKE TESTED` | 8 | the `v0.2.5-dryrun.6` rehearsal published `firecraft-0.2.5-source.tar.gz` (3,176,024 B, sha256 `0cbee184…`) on **draft** 449279892; the archive was downloaded from that draft, re-hashed locally (the digest matches) and extracted on a **second host** (Fedora), where `cargo metadata --locked` succeeds and `cargo build --locked` finished the dev profile in 53.78 s on cargo 1.99.1. The workflow's own step reported `clean-extract + cargo metadata --locked verified`. The archive's only `.claude/` entry is `.allow_config_edits` — the scratch script rehearsal 5 shipped is gone and the archive is 806 B smaller. The naming defect the rehearsal-3 receipt found (the capitalised `FireCraft-0.2.5-source.tar.gz`) is closed | F-9 (the verb is absent from `main`), B-13 (never published to a *release* — a draft is not a publication) | +| 13 | SBOM | CycloneDX 1.5 JSON | `firecraft-0.2.5-sbom.cdx.json` — **not on `main`** | `xtask dist --sbom` over the locked dependency graph; the serial is a deterministic UUIDv8 derived from the locked graph. The verb exists on this branch and on the PR #9 / #12 trees | `SMOKE TESTED` | 8 | the `v0.2.5-dryrun.6` rehearsal published `firecraft-0.2.5-sbom.cdx.json` (369,745 B, sha256 `e50516b2…`) on **draft** 449279892; the workflow's step reported `562 components, 562 locked package(s), schema-valid` and the aggregate gate validated the download. The document is validated against the vendored CycloneDX 1.5 schema **before** it is written and its component inventory is compared with `Cargo.lock`'s own `[[package]]` sections ([decision 0119](../decisions/0119-the-sbom-is-validated-against-the-schema-it-claims.md)). Two independent rehearsals' SBOMs are identical in every field except `metadata.timestamp` — same components, same dependency graph, same serial number | F-9 (the verb is absent from `main`), B-13 (never published to a *release*) | +| 14 | Metadata | checksums / manifest | `dist/native/MANIFEST.sha256` (3 native artifacts) and `dist/web/MANIFEST.sha256` (5 bundle files) | `xtask dist --native` / `dist --web`; `release.yml` additionally writes a release-wide `RELEASE_SHA256.txt` over 7 assets | `PUBLISHED` | 11 | `MANIFEST.sha256` attached to release v0.2.5 (398 B, 3 native files) and embedded in the web ZIP (532 B). Both checked against the published downloads on 2026-11-09 — 3/3 and 5/5 pass. **The release-wide list now exists and verifies**: the `v0.2.5-dryrun.6` rehearsal published `RELEASE_SHA256.txt` (664 B, sha256 `af054b99…`) on **draft** 449279892; downloaded from the draft, its 7 entries match the API digests of the 7 assets it covers — **7 of 7**. The list names each file explicitly because the globs over-match, and every published asset now carries the lowercase canonical name. The native `MANIFEST.sha256` is byte-identical across two independent rehearsals | B-17 (the published release still carries no release-wide list; the list exists only on the rehearsal draft), B-21 (the manifest and the assets it covers were published out of band, so the list is self-consistent but not independently attested) | | 15 | Metadata | provenance / attestation | **nothing on any branch** | no generator, no `gh attestation` step, no in-toto/SLSA material anywhere in the repository | `NOT STARTED` | 0 | nothing exists | B-14 | -| 16 | Metadata | release manifest (JSON) | `dist/release-manifest.json` on `feat/release-manifest` — not on `main`, never published | `xtask dist --manifest` (schema `firecraft.release-manifest/1`); `xtask dist --verify-manifest` re-checks it against the files; `release.yml` generates it after the other assets and publishes it as a ninth release asset | `IMPLEMENTED` | 0 | generator, verifier and unit tests (shape, determinism, rejection paths) exist in `xtask/src/dist.rs`; the aggregate validation checks the manifest against the downloaded assets. **No CI receipt, no published manifest** — [decision 0089](../decisions/0089-release-manifest-shape.md) | B-14 (nothing published), T27-53 (the canonical naming the asset table inherits is still open) | +| 16 | Metadata | release manifest (JSON) | `dist/release-manifest.json` — the workflow now generates it and publishes it as a release asset | `xtask dist --manifest` (schema `firecraft.release-manifest/1`); `xtask dist --verify-manifest` re-checks it against the files; `release.yml` generates it after the other assets and publishes it as a ninth release asset | `SMOKE TESTED` | 8 | the `v0.2.5-dryrun.6` rehearsal published `release-manifest.json` (3,295 B, sha256 `54b8a231…`) on **draft** 449279892; the aggregate gate validated it against the downloaded assets, and the file was re-hashed locally (the digest matches). **It names its tag and its run**: `tag` `v0.2.5-dryrun.6`, `source_sha` `c396b18afba0a…`, `run_url` `…/actions/runs/44188107435`, `run_id` `44188107435`, `built_at` `2026-11-17T22:31:44Z`, 8 artifact entries. Two gaps had to be closed first — the caller now states the tag and `xtask` fails closed when the stated tag does not name the workspace version, and the run URL is reconstructed from `GITHUB_SERVER_URL`, `GITHUB_REPOSITORY` and `GITHUB_RUN_ID` because GitHub Actions exports no `GITHUB_RUN_URL`. Shape: [decision 0089](../decisions/0089-release-manifest-shape.md) | B-14 (nothing is signed and no provenance exists), and the manifest has never been published to a *release* — the draft is not a publication | | 17 | Metadata | signature (Authenticode / Developer ID / GPG) | **nothing**; no signing identity is configured anywhere | no signing step in any workflow; signing remains an operator ceremony | `NOT STARTED` | 0 | nothing exists | B-14 — no artifact can reach rung 9 until this is resolved, so **no artifact can be `CERTIFIED`** | | 18 | Android arm64 | APK/AAB | feasibility report only — [android-feasibility.md](../research/android-feasibility.md) | not started: no native entrypoint, no NDK/SDK CI image | `DEFERRED` | 0 | the report concludes "feasible with bounded work"; the gate decision is recorded in [decision 0121](../decisions/0121-the-android-gate-defers-the-native-port.md) — defer the port, keep Android experimental | no NDK/SDK image, no physical ARM64 device | | 19 | Linux | RPM | no spec, no recipe | — | `DEFERRED` | 0 | [decision 0070](../decisions/0070-rpm-deferred.md), accepted 2026-11-06, re-validated 2026-11-16 (T27-31: no RPM demand on the issue tracker; Fedora-family users are served by the tar.gz) | decision 0070 is not on `main`; RPM is **not** reopened by [decision 0075](../decisions/0075-release-scope-mvp.md) | @@ -72,22 +72,48 @@ | `READY` | 0 | — | | `IN PROGRESS` | 0 | — | | `BLOCKED` | 0 | — | -| `IMPLEMENTED` | 1 | release manifest JSON | -| `BUILD VERIFIED` | 3 | WebOS package, source archive, SBOM | +| `IMPLEMENTED` | 0 | — | +| `BUILD VERIFIED` | 1 | WebOS package | | `INSTALLED` | 0 | — | | `LAUNCHED` | 0 | — | -| `SMOKE TESTED` | 0 | — | +| `SMOKE TESTED` | 3 | source archive, SBOM, release manifest JSON | | `PUBLISHED` | 6 | desktop tar.gz, CLI tar.gz, DEB, web bundle, CLI binaries, checksums/manifest | | `CERTIFIED` | 0 | — | | `DEFERRED` | 4 | Android arm64, Linux RPM, Linux aarch64, Windows ARM64 | | `REJECTED` | 0 | — | -**No artifact has reached `CERTIFIED`.** Five artifacts (desktop tar.gz, -CLI tar.gz, DEB, web bundle, CLI binaries) reach **rung 11** — they were -downloaded from the published release and verified against their -published checksum lists on 2026-11-09, and the CLI binary was launched. -Nothing has reached rung 9: no artifact is signed and no provenance -exists. +**No artifact has reached `CERTIFIED`.** Six artifacts (desktop tar.gz, +CLI tar.gz, DEB, web bundle, CLI binaries, checksums/manifest) reach +**rung 11** — they were downloaded from the published release and +verified against their published checksum lists on 2026-11-09, and the +CLI binary was launched. Nothing has reached rung 9: no artifact is +signed and no provenance exists. + +**The rehearsal path reaches rung 8 and no further.** The +`v0.2.5-dryrun.6` rehearsal (run 44188107435, draft 449279892) +produced all nine artifacts, ran the web bridge smoke green, passed the +DEB's eight-stage package verification in a clean Debian 12 container, +and had three of its assets independently re-verified from the draft +(7/7 checksum entries, 3/3 re-hashes, one clean build on a second +host). A **draft is not a publication**, so those bytes stop at rung 8 +until a release is cut from the tag. The receipt is +[RECEIPT-0.2.5-DRYRUN-6.md](RECEIPT-0.2.5-DRYRUN-6.md); the scope that +this evidence can honestly certify is in +[CERTIFICATION.md](CERTIFICATION.md). + +## Linux platform compatibility + +**What is tested on which Linux distribution is published separately**, +in [LINUX-COMPAT.md](LINUX-COMPAT.md) — it is the authoritative table +for "does this work on my machine", and it lists the untested +combinations explicitly rather than leaving them to be inferred. + +As of 2026-11-17 the matrix has been run across three distro families +and two display servers (Ubuntu 24.04, Debian 12, Fedora 45 × X11 and +Wayland); all six cells pass, and the DEB's eight-stage package +verification passes on both Debian-family cells. The AppImage and +Flatpak rows above remain **never built**, and the RPM row remains +**deferred** — the compat table says so in the same words. ## Build surfaces (what actually exists) @@ -233,11 +259,16 @@ artifacts without it). Tracked by issue #81 / task #158, defect **B-6**. decision is that the site is redeployed from the release bytes. The redeploy itself is gated behind maintainer approval. 5. **The release workflow has never produced a release** (B-4, B-21). - All six published releases were operator-published by hand. + All six published releases were operator-published by hand. The + workflow's own path now runs end to end and produces **drafts** — + `v0.2.5-dryrun.5` and `v0.2.5-dryrun.6`, both green — but a draft is + not a release. 6. **The source archive and the SBOM have never been published** (B-13), - and their verbs do not exist on `main` (F-9). + and their verbs do not exist on `main` (F-9). Both are now published + on the rehearsal **drafts** and independently verified there. 7. **No release-wide checksum file has ever been published** (B-17), - and every published asset reports a null GitHub digest. + and every published asset reports a null GitHub digest. The workflow + writes and publishes one on its drafts, where it verifies 7 of 7. 8. **Nothing is signed and no provenance exists** (B-14), so nothing can be `CERTIFIED`. 9. **RPM is deferred** by [decision 0070](../decisions/0070-rpm-deferred.md) diff --git a/docs/release/CERTIFICATION.md b/docs/release/CERTIFICATION.md new file mode 100644 index 00000000..3019da84 --- /dev/null +++ b/docs/release/CERTIFICATION.md @@ -0,0 +1,290 @@ +# Release certification — FireCraft 0.2.5 + +> What this repository can honestly certify about the 0.2.5 release +> line, what it cannot, and what would have to change before a +> production release is published. +> +> Authority: [decision 0075](../decisions/0075-release-scope-mvp.md) +> (the approved release scope) and +> [ARTIFACT-MATRIX.md](ARTIFACT-MATRIX.md) (the per-artifact evidence +> ledger). Execution ledger: [EXECUTION-BOARD.md](EXECUTION-BOARD.md), +> rows T27-19 and T27-60. + +## What is certified, and what is not + +**Nothing in this repository is `CERTIFIED` in the strict sense, and +this report does not change that.** The evidence ladder in +[PHASE-0-AUDIT.md](PHASE-0-AUDIT.md) §0.3 puts *security/signature +checks completed* at rung 9, and this repository has no signing +identity, no attestation generator and no provenance material. Every +artifact here therefore stops at rung 8 (functional smoke completed) on +the rehearsal path, and at rung 11 (downloaded published artifact +independently verified) only for the assets that were published out of +band on 2026-11-09. + +What this report *does* certify is narrower and real: + +1. The release workflow **runs end to end** and produces the approved + artifact set with a fail-closed aggregate gate. +2. The **artifacts themselves** are real, installable, launchable and + functionally smoke-tested on the platforms named below, and their + bytes are reproducible. +3. The **scope** is recorded and authorized — decision 0075 is + accepted and names the approver and the date. +4. The **publication** step is *not* authorized and has not been + taken; the only publication that exists is the out-of-band + `v0.2.5` release, and it carries three open defects. + +## The release identity + +| Fact | Value | Where it was read | +|---|---|---| +| Workspace version | `0.2.5` | `cargo metadata` (the workspace's `xtask` package version) | +| Source SHA (rehearsal head) | `c396b18a` | the rehearsal run's `head_sha` | +| Rehearsal tag | `v0.2.5-dryrun.6` | the draft release's `tag_name` | +| Rehearsal workflow run | `44188107435` | `GET /actions/runs/44188107435` | +| Draft release | id **449279892**, `draft: true`, `prerelease: true`, title `TEST 0.2.5-dryrun.6` | `GET /releases/449279892` | +| Rehearsal tag shape | **lightweight** — `GET /git/ref/tags/v0.2.5-dryrun.6` reports an object of type `commit`, and `GET /git/tags/…` answers 404, so the tag carries no message of its own | same | +| Release mode | rehearsal | `scripts/classify_release_mode.sh` — `vX.Y.Z-dryrun.N` is a rehearsal, never a production release | +| Release state | **draft, unpublished** | `GET /releases/` → `draft: true`, `published_at: null` | +| Toolchain | rustc / cargo **1.99.1** | the workflows' `toolchain:` pin (decision 0102) | +| Native build image | `rust:1.99-slim-bookworm` (Debian 12) | `.github/workflows/release.yml` | +| CI image | `rust:1.99.1-slim-bookworm` | `infrastructure/docker/ci-rust.Dockerfile` | +| GitHub runner | `ubuntu-24.04` | the workflow's `runs-on` | +| GitForge runner | the Fedora host | the CI job store's `runner_id` | + +The rehearsal receipt with the run-by-run detail is +[RECEIPT-0.2.5-DRYRUN-6.md](RECEIPT-0.2.5-DRYRUN-6.md). + +## The approved scope and its matrix + +[decision 0075](../decisions/0075-release-scope-mvp.md) approved nine +items as the release scope. This is where each one stands: + +| # | Approved item | Artifact | Rung | Evidence | +|---|---|---|---|---| +| 1 | Web / WASM bundle | `firecraft-web-0.2.5.zip` | 8 | built by the workflow; the bridge smoke ran green in headless Chromium against the built bundle | +| 2 | Linux x86_64 desktop archive | `firecraft-0.2.5-linux-x86_64.tar.gz` | 8 | installed and launched on six cells; the bytes are the published rehearsal bytes | +| 3 | Linux x86_64 DEB | `firecraft_0.2.5_amd64.deb` | 8 | installed with `apt` on Ubuntu 24.04 and Debian 12; the repository's own eight-stage verification passes | +| 4 | Linux x86_64 CLI archive | `firecraft-cli-0.2.5-linux-x86_64.tar.gz` | 8 | extracted and run on six cells; `firecraft-cli --version` prints the stamped version | +| 5 | Source archive | `firecraft-0.2.5-source.tar.gz` | 8 | clean extraction on a second host runs `cargo metadata --locked` and `cargo build --locked` to completion | +| 6 | Checksums | `RELEASE_SHA256.txt` | 8 | 7 of 7 entries match the API's own digests for the assets it covers | +| 7 | SBOM | `firecraft-0.2.5-sbom.cdx.json` | 8 | validated against the vendored CycloneDX 1.5 schema and compared component-by-component with `Cargo.lock` | +| 8 | Verified production web deployment | — | **blocked** | the live deployment serves a bundle that matches no release (see below) | +| 9 | Release certification | this document | — | exists; it certifies nothing beyond the evidence it lists | + +**Item 8 is not satisfied.** The approved scope asks for a *verified* +production deployment, and the deployment cannot be verified: see +[Deployment reconciliation](#deployment-reconciliation). This is the +one approved-scope item that is blocked, and it blocks the release +cut on its own evidence — not on authorization. + +## The artifacts, as published on the rehearsal draft + +Every asset is `state: uploaded`; the digest is the GitHub API's own +`sha256:` field for the asset bytes, not a digest this repository +computed. + +| Asset | Bytes | sha256 | Runner / toolchain | +|---|---|---|---| +| `firecraft-0.2.5-linux-x86_64.tar.gz` | 24,389,908 | `cb55a52507477c9af327c35be8bb0e6d7dfc16a767e34bfde1aab0d92f346a3d` | ubuntu-24.04 · `rust:1.99-slim-bookworm` (Debian 12) · rustc 1.99.1 | +| `firecraft-cli-0.2.5-linux-x86_64.tar.gz` | 8,158,124 | `953633e40c2021858e7001f9691a47ad4ed4491ec752b38ccd716f8e99339093` | same | +| `firecraft_0.2.5_amd64.deb` | 16,228,104 | `7928369626375dfca80516551870d0ab7c98b3f3021ff5c82fc36a86dc39fa42` | same | +| `firecraft-web-0.2.5.zip` | 5,853,417 | `898be851f2e1abac03e0e01a58a3d5f593e74dd44acd572f0d47c4cbfef9d3c4` | same (wasm 15,780,814 B) | +| `firecraft-0.2.5-source.tar.gz` | 3,176,024 | `0cbee1846cb24211fe3be899d286dd1917fd2464d2219f754ae9df7744d5f523` | same | +| `firecraft-0.2.5-sbom.cdx.json` | 369,745 | `e50516b21fcb5254824a262d73649d746b971a27b1d4a4ffcabe554f0b703174` | same (562 components, 562 locked packages) | +| `MANIFEST.sha256` | 398 | `6a8692ed6806fe9074e8f29762bbbbaabd78ace8bba3e1c5bd5702ead4a5a101` | same | +| `release-manifest.json` | 3,295 | `54b8a231e2d4ff7e155915867f8d085dcfb86f944815032bd67e06ce2277ee57` | same (schema `firecraft.release-manifest/1`) | +| `RELEASE_SHA256.txt` | 664 | `af054b99c8c3b864de127597f23bf0f133b4d65f6504de70ba8ede11a61e8381` | same | + +All nine are `state: uploaded` on draft release **449279892** +(`tag_name` `v0.2.5-dryrun.6`, `draft: true`, `prerelease: true`, +`published_at: null`). + +## Install, launch and smoke evidence + +The evidence is the **six-cell distro × display matrix**, run on the +Fedora host on 2026-11-17 against the repository's own tracked scripts +(`scripts/release-matrix.sh` and `packaging/linux/leg-*.sh` at +`8f9a602c`). + +| Distribution | Version | DEB | desktop tar.gz | CLI tar.gz | X11 | Wayland | +|---|---|---|---|---|---|---| +| Ubuntu | 24.04.5 LTS | installed + 8 stages | launched | ran | pass | pass | +| Debian | 12 (bookworm) | installed + 8 stages | launched | ran | pass | pass | +| Fedora | 45 | n/a — format | launched | ran | pass | pass | + +The Debian-family cells install the package with `apt` — so apt itself +resolves the declared dependencies — and then run the repository's own +`packaging/deb/verify-install.sh`: metadata, desktop entry (accepted by +`desktop-file-validate`), icon, CLI, CLI document round trip (a 97 B +PNG with the signature `89504e470d0a1a0a`), GUI launch under Xvfb +(stayed up for the 25 s launch window, exit 124), uninstall and +reinstall. The Wayland cells run weston's headless backend and launch +the binary with `WAYLAND_DISPLAY=wayland-0`. The full statement of what +each cell did, and of what no cell did, is +[LINUX-COMPAT.md](LINUX-COMPAT.md). + +The web bundle's smoke is the bridge round trip — open → apply → +`exportPng` → `exportDocument` — run by `xtask dist --web` in headless +Chromium. It is **not** skipped: the release workflow installs headless +Chrome and fails closed without it, and the environment variable that +would skip it appears nowhere in the CI configuration. + +## Signature, SBOM and provenance status + +| Property | Status | Consequence | +|---|---|---| +| Artifact signatures | **none** — no Authenticode, Developer ID or GPG signing step exists in any workflow, and no signing identity is configured | no artifact can reach rung 9; no artifact is `CERTIFIED` | +| Provenance / attestation | **none** — no in-toto, SLSA, `cosign` or `gh attestation` machinery exists in the repository (board T27-56, blocker B-14) | the release set is not bound to a source commit by anything a third party can verify | +| SBOM | **present** — CycloneDX 1.5 JSON, validated against the vendored schema and compared component-by-component with `Cargo.lock`; the comparison names both sides of any difference | the *dependency inventory* is attested; the *artifact* is not | +| Release manifest | **present** — `release-manifest.json`, schema `firecraft.release-manifest/1`, carrying the version, the tag, the source SHA, the run URL, the timestamp and one entry per artifact with its size, digest, format and target triple | the release describes itself; a third party cannot verify the description without the digests being independently attested | + +The gap between "the release describes itself" and "the release is +attested" is exactly what T27-56 and blocker B-14 describe, and it is +the reason this report stops where it stops. + +## Deployment reconciliation + +The comparison is a tracked tool and a machine-readable report: +`scripts/reconcile_deployment.py` fetches the live deployment's +`MANIFEST.sha256` and every file it names, hashes the bytes it +actually received, and compares them file by file with the release +bundle's manifest. The report is +[DEPLOYMENT-RECONCILIATION.json](DEPLOYMENT-RECONCILIATION.json), and +its result is **`differences`**. + +- The live deployment serves bundle `2274ac8c2925b61d`. +- The `v0.2.5` release bundle carries bundle `f9dbe2f14305b2d9`. +- Each asset name appears only on its own side of the inventory. +- `appinfo.json` matches byte for byte (`7d039c36…`); `index.html` + and `host.html` differ only by the substituted asset names, their + integrity digests and the Pages analytics beacon. + +**The live deployment corresponds to no identifiable release.** The +deploy step in `release.yml` is commented out and no +`CLOUDFLARE_API_TOKEN` is provisioned (blockers B-11, B-12), so the +deployment cannot be advanced from the release pipeline and cannot be +reconciled from it either. Approved-scope item 8 stays blocked. + +**Re-verified 2026-11-17.** The comparison was run again against the +live site: the deployment's `MANIFEST.sha256` digest is unchanged at +`61e3aeda5e8dcc340725a52c6a421422198f29b9ee8c1ed9a91832f372985261`, +and the comparison against the rehearsal's web bundle reports eight +differences — `appinfo.json` matches, and every other file differs. +The live bundle has not moved. + +## CI links and the state of the gates + +| Gate | What it proves | Where | +|---|---|---| +| `cargo fmt --all -- --check` | the tree is rustfmt-clean | GitHub `fmt`; GitForge `fmt-layers` | +| `cargo clippy --workspace --all-targets -- -D warnings` | no lint, in any target including the tests | GitHub `gate`; GitForge `clippy-test` | +| workspace tests | the suite passes | GitHub `gate`; GitForge `clippy-test` | +| `xtask layers` | the crate dependency layers hold | GitHub `gate`; GitForge `layers` | +| Aegis secret scan | no new secret finding against the content-keyed baseline | GitHub `gate`; GitForge `aegis` | +| the release ledger | the board and its tickets agree | GitHub `gate`; GitForge `release-ledger` | +| the aggregate release gate | every required artifact is present, non-empty, correctly named, digest-matching and platform-covering | `scripts/validate_release_aggregate.sh`, run by the release workflow | + +The rehearsal also exercised the release workflow's own pin and mode +gates: `scripts/validate_workflow_pins.py` (every third-party action +pinned to a full commit SHA), `scripts/classify_release_mode.sh` (a +manual run may only rehearse; a rehearsal is a prerelease titled +`TEST ` with a do-not-publish banner), and +`scripts/validate_release_aggregate.sh` (fails closed on a missing, +empty, extra, misnamed or digest-mismatched artifact, and on a skipped +required platform). + +## Limitations + +- **Nothing is signed or attested.** See the table above; this is the + load-bearing limitation of the whole report. +- **The rehearsal is a draft.** Its assets are published to a draft + release, not to a published release, and its tag is disposable. The + evidence in this report is about *the rehearsal's bytes*, which are + the bytes the workflow produces — not about a published release. +- **The production `v0.2.5` assets are not these bytes.** They were + published out of band on 2026-11-09 (blocker B-21), they are smaller + than the workflow's rebuilds, and they carry the three open defects + below. The matrix in this report tests the *rehearsal's* bytes. +- **The matrix ran in containers**, with software rendering, under Xvfb + and weston's headless backend. No cell used a real GPU, a real + compositor or a real desktop session; desktop integration was + checked at the package level only. +- **The DEB's upgrade path was not exercised** — the eight-stage + verification needs a previous package to upgrade from, and none was + supplied. +- **No Windows, macOS, AppImage, Flatpak, Android, ARM64 or RPM + artifact exists.** Those are outside the approved scope + ([decision 0075](../decisions/0075-release-scope-mvp.md), extended + campaign) and are not claimed here. +- **The web deployment is unreconciled and the live site's bytes match + no release.** + +## Unsupported targets + +Windows (portable ZIP and Inno Setup installer), macOS (`.app` and +DMG), Linux AppImage, Linux Flatpak, Linux RPM, Linux aarch64, +Windows ARM64 and Android are **not supported by this release**. Each +is either outside the approved scope or inside the extended campaign +with its own acceptance evidence outstanding. Nothing in this +repository may advertise them as supported before that evidence +exists. + +## Residual risk + +| Risk | Severity | Why it is still open | +|---|---|---| +| A consumer cannot verify that a downloaded artifact is the one this repository built | **High** | no signature and no provenance exist (T27-56, B-14) | +| The live web deployment is not a release | **High** | the deploy step is disabled and no token is provisioned (B-11, B-12); the deployed bundle matches no release | +| The published `v0.2.5` DEB installs cleanly and then fails to start on older distributions | **High** | the published asset under-declares its glibc floor (B-7); the workflow's rebuild fixes it, but the published asset is out of band and unreplaced | +| The published `v0.2.5` DEB is not a desktop application | **Medium** | it carries no desktop entry and no icon (B-24); the workflow's rebuild carries both | +| The published `v0.2.5` DEB declares only `libc6` | **Medium** | it omits the X11/GL runtime stack the binary needs (issue #39); the workflow's rebuild declares the full set | +| The published `v0.2.5` assets are not reproducible from the tree | **Medium** | they were published out of band (B-21); the workflow's own rebuilds *are* byte-identical to each other across rehearsals | +| A release cut is published from a tree whose gates were not run on that exact SHA | **Low** | mitigated by requiring the GitForge receipt for the same candidate SHA before any merge or publication | + +## Independent verification results + +These are checks run **against the published bytes**, not against the +build that produced them. + +| Check | Result | +|---|---| +| `RELEASE_SHA256.txt` downloaded from the draft and compared with the API's own digests | **7 of 7 match** (the list does not cover itself or the manifest, so it holds seven entries for nine assets) | +| The source archive, `RELEASE_SHA256.txt` and `release-manifest.json` downloaded and re-hashed locally | **3 of 3 match** the API's digests byte for byte | +| The source archive extracted, then `cargo metadata --locked --offline` | exit **0** | +| The source archive extracted on a second host (Fedora), then `cargo metadata --locked` and `cargo build --locked` | `METADATA OK`, then `Finished dev profile … in 53.78s` on cargo 1.99.1 | +| The source archive's contents | the only `.claude/` entry is `.allow_config_edits` — no scratch helper is published | +| The native artifacts of two independent rehearsals compared byte for byte | **identical** — the desktop tarball, the CLI tarball, the DEB and `MANIFEST.sha256` have the same digests in two separate runs | +| The web bundle's payload of two independent rehearsals compared file by file | **identical** — all seven files hash the same; only the ZIP container's timestamps differ | +| The SBOM of two independent rehearsals compared field by field | identical in every field except `metadata.timestamp`; the 562 components, the dependency graph and the serial number match | +| The six-cell matrix run against the published asset digests | all six cells pass; the artifacts under test are the published bytes, not a local rebuild | + +## What publication would require + +1. A maintainer decision on the live web deployment: either provision + `CLOUDFLARE_API_TOKEN` and re-enable the deploy step so that + approved-scope item 8 can be satisfied, or record that the release + is published without a verified deployment. +2. A signing identity and a provenance generator, so that the release + can reach rung 9 and its artifacts can be `CERTIFIED` (T27-56). +3. The maintainer's written publication authorization (T27-61, + blocker B-04). Nothing in this repository may publish a production + release without it; the rehearsals above are what that authorization + would replace. +4. A decision on the out-of-band `v0.2.5` assets — re-cut them from + the workflow, or record them as superseded. + +## Related documents + +- [ARTIFACT-MATRIX.md](ARTIFACT-MATRIX.md) — one row per artifact, + with its evidence state and blockers. +- [EXECUTION-BOARD.md](EXECUTION-BOARD.md) — the per-unit ledger. +- [LINUX-COMPAT.md](LINUX-COMPAT.md) — what is *tested* on Linux, and + what is only built or deferred. +- [RECEIPT-0.2.5-DRYRUN-6.md](RECEIPT-0.2.5-DRYRUN-6.md) — the + rehearsal receipt behind this report. +- [PHASE-0-AUDIT.md](PHASE-0-AUDIT.md) — the evidence baseline, the + evidence ladder and blockers B-1…B-21. +- [../decisions/0075-release-scope-mvp.md](../decisions/0075-release-scope-mvp.md) + — the approved release scope. diff --git a/docs/release/EXECUTION-BOARD.md b/docs/release/EXECUTION-BOARD.md index 73782360..4ebf0347 100644 --- a/docs/release/EXECUTION-BOARD.md +++ b/docs/release/EXECUTION-BOARD.md @@ -186,8 +186,8 @@ condition/verification. **RP** is the release-plan work package the row belongs | ID | Title | Pri | Status | Depends on | Acceptance criteria | Verification | Evidence | Blockers | Done | |---|---|---|---|---|---|---|---|---|---| | T27-20 | Prove the DEB glibc-floor check fails on a too-low floor and passes on an adequate one | P0 | IN PROGRESS | T27-06 | The `objdump`-based floor check is demonstrated to reject a too-low declaration and accept an adequate one | `cargo run -p xtask -- dist --native` | `b63a040` added `verify_deb_glibc_floor`; control now floors `libc6 (>= 2.35)`; decision 0091 adds the run-time library set (`libx11-6`, `libxcursor1`, `libxrandr2`, `libxi6`, `libxkbcommon0`, `libxkbcommon-x11-0`, `libgl1`, `libgcc-s1`) with a build gate that refuses a control file that drops one | — | — | -| T27-21 | Build the DEB from the candidate commit, inspect its metadata, install it clean and run a CLI round trip | P0 | IN PROGRESS | T27-20 | DEB builds from the candidate commit; `dpkg-deb --info/--contents` agree with the metadata; clean Debian 12 install plus CLI import→export round trip passes | `dpkg-deb --info dist/native/firecraft_0.2.5_amd64.deb && apt install ./firecraft_0.2.5_amd64.deb` | Prior clean-container smoke passed 2026-11-06 (matrix); the pipeline's DEB install verification ran stages 1–4b green on `4273471` (run `85502100`) and stage 5 (GUI launch) failed on the undeclared `libxkbcommon-x11-0` — decision 0091 declares the run-time set and the CI installs exactly it | Clean Debian 12 environment | — | -| T27-22 | Verify DEB reinstall/remove/purge and ship a desktop entry and icon | P0 | IMPLEMENTED | T27-21 | Reinstall, remove and purge succeed with no user data lost; the package ships a desktop entry and an icon | `dpkg -l; dpkg -P; dpkg-deb --contents … \| grep -E 'desktop\|icons'` | The row's premise is fixed in the tree. `build_deb` stages the desktop entry and the icon from the in-tree originals — `usr/share/applications/firecraft.desktop` and `usr/share/icons/hicolor/256x256/apps/firecraft.png` (`xtask/src/dist.rs:973`) — and the gate reads the built package back with `dpkg-deb --contents` and fails closed when either is missing. **The exercise against the fixed package has not been run**: [`packaging/deb/verify-install.sh`](https://github.com/aliasfoxkde/FireCraft/blob/main/packaging/deb/verify-install.sh) carries the eight stages (install, desktop entry, icon, CLI, CLI document round trip, GUI launch, upgrade, uninstall, reinstall), and the recorded negative control is the *published* v0.2.3 package, which installs cleanly and then fails stage 2 — no desktop entry (audit §11.6). The defect is fixed; the reinstall / remove / purge receipt against a package that carries the desktop entry does not exist. | the clean-container eight-stage run against a package built from the fixed tree (the CI image and a built .deb) | 2026-11-11: the desktop entry and the icon landed in the DEB payload (PR #107); the clean-container receipt is outstanding | +| T27-21 | Build the DEB from the candidate commit, inspect its metadata, install it clean and run a CLI round trip | P0 | SMOKE TESTED | T27-20 | DEB builds from the candidate commit; `dpkg-deb --info/--contents` agree with the metadata; clean Debian 12 install plus CLI import→export round trip passes | `dpkg-deb --info dist/native/firecraft_0.2.5_amd64.deb && apt install ./firecraft_0.2.5_amd64.deb` | the DEB builds from the candidate commit inside the release workflow's Debian 12 container (`rust:1.99-slim-bookworm`), `dpkg-deb` reads the package back and the gate fails closed when the desktop entry or the icon is missing from it. The workflow's own package verification then ran against that package in a clean container and **all eight stages passed** (run 44188107435, `build artifacts` job, 2026-11-17T22:24:48Z → 22:25:14Z): install, desktop entry, icon, CLI, CLI document round trip, GUI launch under Xvfb, **upgrade**, uninstall and reinstall. The separate distro matrix installs the same bytes with `apt` on Ubuntu 24.04 and Debian 12 and passes stages 1–5, 7 and 8 ([LINUX-COMPAT.md](../release/LINUX-COMPAT.md)). The upgrade stage is exercised by the workflow's verification, which the matrix does not supply. | — | 2026-11-17: the workflow's eight-stage package verification passed on the rehearsal bytes | +| T27-22 | Verify DEB reinstall/remove/purge and ship a desktop entry and icon | P0 | IMPLEMENTED | T27-21 | Reinstall, remove and purge succeed with no user data lost; the package ships a desktop entry and an icon | `dpkg -l; dpkg -P; dpkg-deb --contents … \| grep -E 'desktop\|icons'` | The row's premise is fixed in the tree. `build_deb` stages the desktop entry and the icon from the in-tree originals — `usr/share/applications/firecraft.desktop` and `usr/share/icons/hicolor/256x256/apps/firecraft.png` (`xtask/src/dist.rs:973`) — and the gate reads the built package back with `dpkg-deb --contents` and fails closed when either is missing. **The exercise against the fixed package has not been run**: [`packaging/deb/verify-install.sh`](https://github.com/aliasfoxkde/FireCraft/blob/main/packaging/deb/verify-install.sh) carries the eight stages (install, desktop entry, icon, CLI, CLI document round trip, GUI launch, upgrade, uninstall, reinstall), and the recorded negative control is the *published* v0.2.3 package, which installs cleanly and then fails stage 2 — no desktop entry (audit §11.6). The defect is fixed. **The clean-container receipt against a package that carries the desktop entry now exists** (2026-11-17): the eight-stage run passes on Ubuntu 24.04.5 LTS and Debian 12 — stage 2 `PASS: desktop entry installs and names firecraft / firecraft`, `PASS: desktop-file-validate accepts the entry`; stage 3 `PASS: icon installed at /usr/share/icons/hicolor/256x256/apps/firecraft.png`; stage 7 `PASS: removal took the binaries, the desktop entry and the icon with it`; stage 8 `PASS: reinstall restored the payload and the CLI runs`. The row stays `IMPLEMENTED` rather than closing, because **purge is not separately exercised** — stage 7 is an uninstall, and no `dpkg -P` run is in the receipt. | purge (`dpkg -P`) is not exercised by the eight-stage script; a purge receipt is still outstanding | 2026-11-17: the clean-container eight-stage receipt exists on Ubuntu 24.04 and Debian 12 for a package that carries the desktop entry and the icon | | T27-26 | Build an actual AppImage with linuxdeploy | P1 | NOT STARTED | T27-06 | A real `.AppImage` is produced by a continuous linuxdeploy build (not extract-and-run) | `packaging/build-appimage.sh && ls -l dist/firecraft-*-linux-x86_64.AppImage` | `a4d74a0` fixed the missing `Icon=` blocker; a full re-run is still pending | linuxdeploy binary on the build host | — | | T27-27 | Inspect the built AppImage and launch it both ways | P1 | NOT STARTED | T27-26 | The image carries a desktop entry with `Icon=firecraft`, the icon file, executable bits and bundled libraries; it launches under FUSE and via `--appimage-extract-and-run` | `./firecraft-*-linux-x86_64.AppImage --appimage-extract-and-run` | — | FUSE availability on the host | — | | T27-28 | Functionally smoke the AppImage on a clean Linux host and record its checksum | P1 | NOT STARTED | T27-27 | Open → edit → save/export succeeds from the AppImage on a clean host; SHA-256 and receipt recorded | `sha256sum dist/firecraft-*-linux-x86_64.AppImage` | — | Clean host | — | @@ -195,8 +195,8 @@ condition/verification. **RP** is the release-plan work package the row belongs | T27-30 | Install the Flatpak into a test installation and exercise it under the sandbox | P1 | NOT STARTED | T27-29 | The app installs into a test installation and runs via `flatpak run` (no unsupported `--repo`); document-portal open/save/export works; removal works; permissions are minimized and each is documented | `flatpak run dev.firecraft.FireCraft` | the manifest declares `--device=dri --share=ipc --socket=x11 --socket=wayland --socket=fallback-x11 --filesystem=xdg-documents --talk-name=org.freedesktop.portal.Desktop`; none of it has been exercised — the bundle has never been built (T27-29). | a Flatpak SDK/runtime host | — | | T27-31 | Re-validate the RPM deferral against current demand | P2 | IMPLEMENTED | T27-03 | Decision 0070 is re-checked and either retained or formally reinstated with a spec and recipe | `grep -n 'Status' docs/decisions/0070-rpm-deferred.md` | The deferral is re-checked against current demand and **retained**: no issue on the tracker asks for an RPM package (the only RPM mentions are this ledger's own rows), Fedora-family users are served by the tar.gz, and T27-33's second-distro-family check is a DEB/tar.gz exercise. The dated re-validation is recorded in [0070](../decisions/0070-rpm-deferred.md) and the matrix's RPM row cites it | — | 2026-11-16 | | T27-32 | Assess a Linux aarch64 target | P2 | IMPLEMENTED | T27-04 | A written assessment covers the aarch64 toolchain, native dependencies and the packaging path | `rustup target list --installed` | The assessment is the ticket itself ([`docs/tasks/T27-32-linux-aarch64-assessment.md`](https://github.com/aliasfoxkde/FireCraft/blob/main/docs/tasks/T27-32-linux-aarch64-assessment.md)): the toolchain (Rust's aarch64 standard library is installed, but the host has no aarch64 C toolchain and the CI image carries only the wasm target), the native stack (the default build is pure Rust — X11, Wayland and OpenGL are all loaded at run time, and the one C library, `libwebp-sys`, sits behind a non-default feature), the packaging path (the x86_64-pinned artifact names, the DEB's `Architecture` field, and the AppImage and Flatpak output names), the CI cost (one cross-build job plus one image revision), and the verdict **feasible-later**: no code port is needed, but the artifact cannot rise above a build receipt without an ARM64 test environment. Nothing has been cross-built or run on aarch64. | — | 2026-11-16: assessment written | -| T27-33 | Test the DEB and tar.gz on a second distro family and check X11/Wayland, desktop integration and fonts | P1 | NOT STARTED | T27-21 | The DEB and tar.gz install and launch on Ubuntu and on a Fedora-family image; X11 and Wayland, desktop integration, and font/resource loading are each checked | `docker run bash -lc 'apt install ./firecraft_*.deb && firecraft --version'` | only Debian 12 has been exercised for the DEB — the rehearsal receipt `docs/release/RECEIPT-0.2.5-DRYRUN-3.md` records the Debian-family DEB build; no Fedora-family image exists on this fleet. | no Fedora-family image is available locally | — | -| T27-34 | Publish a Linux platform compatibility table | P1 | NOT STARTED | T27-33 | A published table lists every tested distro/arch and every untested gap | `test -f docs/release/LINUX-COMPAT.md` | none yet. | T27-33 | — | +| T27-33 | Test the DEB and tar.gz on a second distro family and check X11/Wayland, desktop integration and fonts | P1 | SMOKE TESTED | T27-21 | The DEB and tar.gz install and launch on Ubuntu and on a Fedora-family image; X11 and Wayland, desktop integration, and font/resource loading are each checked | `scripts/release-matrix.sh dist/native` | Six cells, all passing, on 2026-11-17: Ubuntu 24.04.5 LTS, Debian 12 (bookworm) and Fedora 45, each under X11 and under Wayland. The DEB installs with apt on both Debian-family cells and passes the repository's own eight-stage package verification — metadata, desktop entry accepted by `desktop-file-validate`, icon installed, CLI, CLI document round trip (`firecraft-cli export` writes a 97 B PNG with the signature `89504e470d0a1a0a`), GUI launch under Xvfb with exit 124, uninstall and reinstall. The desktop tar.gz launches on all six cells and `fc-list` enumerates fonts on every cell. The harness is in-tree: `packaging/linux/leg-debian-family.sh`, `packaging/linux/leg-fedora-family.sh`, `packaging/linux/leg-wayland.sh` and [`scripts/release-matrix.sh`](https://github.com/aliasfoxkde/FireCraft/blob/main/scripts/release-matrix.sh). The Fedora cells record that a `.deb` is not that family's package format, so the tar.gz is the artifact under test there. The results are published in [LINUX-COMPAT.md](LINUX-COMPAT.md). | — | 2026-11-17: six-cell matrix green on Ubuntu 24.04, Debian 12 and Fedora 45 | +| T27-34 | Publish a Linux platform compatibility table | P1 | IMPLEMENTED | T27-33 | A published table lists every tested distro/arch and every untested gap | `test -f docs/release/LINUX-COMPAT.md` | [LINUX-COMPAT.md](LINUX-COMPAT.md) is published and linked from `docs/release/ARTIFACT-MATRIX.md`. It lists the three tested distributions with their versions and architecture, the artifact under test in each cell with its bytes and SHA-256, the formats with no tested row (AppImage and Flatpak — never built; RPM — deferred), and the untested combinations explicitly: every other distribution and release, every architecture but x86_64, real GPUs and vendor drivers, real desktop sessions, session-level desktop integration, the DEB's upgrade path, and the published v0.2.5 assets. | — | 2026-11-17: table published and linked | ### RP-02 — Windows packaging and verification @@ -231,20 +231,20 @@ condition/verification. **RP** is the release-plan work package the row belongs | ID | Title | Pri | Status | Depends on | Acceptance criteria | Verification | Evidence | Blockers | Done | |---|---|---|---|---|---|---|---|---|---| | T27-48 | Build the web bundle from the exact tag SHA and record its manifest | P0 | BUILD VERIFIED | T27-06 | The bundle builds from the tag SHA, the WASM is optimized and under the 21.5 MB ceiling, the version is stamped, and a manifest plus per-file hashes are recorded | `cargo run -p xtask -- dist --web && (cd dist/web && sha256sum -c MANIFEST.sha256)` | The bundle was built from the tag's commit and published with its manifest: the rehearsal's `build artifacts` job (run 42937560976, head SHA `384687f`) ran `dist --web` successfully and published `firecraft-web-0.2.5-dryrun.3.zip` (7,394,859 B, sha256 `57054d06…`) with `MANIFEST.sha256` (425 B, sha256 `e71ce485…`); the release-wide list covers the ZIP and the manifest verifies with `sha256sum -c`. Receipt: [`docs/release/RECEIPT-0.2.5-DRYRUN-3.md`](https://github.com/aliasfoxkde/FireCraft/blob/main/docs/release/RECEIPT-0.2.5-DRYRUN-3.md). The WASM is stamped and inside the budget — the GitForge release job for `fcfe6db1` logged `wasm 15573974 B (warning 25000000, hard ceiling 50000000)`, 5 files. | — | 2026-11-12: the rehearsal published the web bundle and its manifest | -| T27-49 | Run the bridge smoke against the release bytes with no skip | P0 | BUILD VERIFIED | T27-48 | The release bytes complete an open → apply → exportPng → exportDocument round trip in headless Chromium, with the smoke not skipped | `cargo run -p xtask -- dist --web` (bridge smoke stage) | The smoke runs and is not skipped. `dist --web` calls `smoke_bridge` ([`xtask/src/dist.rs`](https://github.com/aliasfoxkde/FireCraft/blob/main/xtask/src/dist.rs), line 65), the release workflow installs headless Chrome and fails closed without it (`.github/workflows/release.yml:106`), and `FIRECRAFT_DIST_SKIP_SMOKE` appears nowhere in `.gitforge.yml` or the GitHub workflows — only as a comment at `xtask/src/dist.rs:52`. The rehearsal's `build artifacts` job (run 42937560976) therefore ran the smoke against the tagged bundle and succeeded, and the GitForge release job logs `… smoke green` and `[xtask] dist --web ok` — most recently job `ecb488d1-8965-443f-8309-71405883d754` for commit `fcfe6db1`, finished 2026-11-12T06:18:54Z; twelve release jobs carry that line. | — | 2026-11-12: the bridge smoke ran green against the built bundle | -| T27-50 | Reconcile the production Pages deployment against the release bundle | P0 | IMPLEMENTED | T27-48 | The Pages project, deployment id, deployed asset inventory and per-file hashes are compared to the tagged bundle, and the result is written as a machine-readable report | `curl -s https://firecraft-web.pages.dev/MANIFEST.sha256` | the comparison is a tracked tool and a machine-readable report. [`scripts/reconcile_deployment.py`](https://github.com/aliasfoxkde/FireCraft/blob/main/scripts/reconcile_deployment.py) fetches the deployment's `MANIFEST.sha256` and every file it names, hashes the bytes it actually received, and compares them file by file with the bundle's manifest; [`scripts/reconcile_deployment-selftest.py`](https://github.com/aliasfoxkde/FireCraft/blob/main/scripts/reconcile_deployment-selftest.py) covers the four ways the two can disagree — identical, changed bytes, a deployment-only file, and a bundle file the deployment does not serve. The report against the live deployment is [`docs/release/DEPLOYMENT-RECONCILIATION.json`](https://github.com/aliasfoxkde/FireCraft/blob/main/docs/release/DEPLOYMENT-RECONCILIATION.json) and its result is `differences`: the live deployment serves bundle `2274ac8c2925b61d` while the v0.2.5 release asset carries `f9dbe2f14305b2d9`, each asset name appearing only on its own side of the inventory; `appinfo.json` matches, and `index.html` and `host.html` differ only by the substituted asset names, their integrity digests and the Pages analytics beacon. This confirms the earlier hand comparison (PR #127, `8d5c442`) in machine-readable form: the live deployment corresponds to no identifiable release (blockers B-6, B-11, B-12). | no `CLOUDFLARE_API_TOKEN`; the deploy step in `release.yml` is commented out — a deployment that names a commit stays maintainer-gated | 2026-11-16: the comparison landed as a tracked tool and a machine-readable report | -| T27-51 | Build and verify the standalone CLI archive | P0 | IN PROGRESS | T27-06 | The CLI archive builds from the candidate commit; architecture, `--version`/`--help`, an import→export round trip, clean extraction and the checksum all verify | `firecraft-cli --version && firecraft-cli import … && firecraft-cli export …` | the matrix records the CLI tar.gz as built and manifested; rehearsal `v0.2.5-dryrun.3` (run 42937560976) published `firecraft-cli-0.2.5-dryrun.3-linux-x86_64.tar.gz` (8,109,224 B, sha256 `f646feea…`) as a draft asset and the extracted binary launches. **The archive cannot attest its own build**: the CLI had no `--version`, so the receipt had nothing to say which build it was. `firecraft-cli --version` now prints the stamped version — the archive must be rebuilt for this row to close. | — | 2026-11-11: rehearsal receipt written; the rebuilt archive is outstanding | -| T27-52 | Generate the source archive from the exact tagged commit and clean-build it | P0 | IN PROGRESS | T27-06 | The archive is generated from the tag SHA, contains the sources, licenses and build inputs, excludes build outputs and secrets, and a clean extraction runs `cargo metadata --locked` and `cargo build --locked` | `cargo run -p xtask -- dist --source` | the generator is implemented and wired into `release.yml`; **exercised — and it failed**. The rehearsal's archive (`FireCraft-0.2.5-dryrun.3-source.tar.gz`, asset 733682287) was downloaded from the draft and extracted: `cargo metadata --locked` exits **101** ("cannot update the lock file … because --locked was passed to prevent this"). The manifests are stamped; the lock's dependency *references* and the path dependency's version requirement were not. Fixed by the `xtask stamp` verb and the `xtask stamps` drift gate; the row closes when a rebuilt archive clean-builds. | — | 2026-11-11: defect reproduced from the published bytes; fix implemented, verification outstanding | +| T27-49 | Run the bridge smoke against the release bytes with no skip | P0 | SMOKE TESTED | T27-48 | The release bytes complete an open → apply → exportPng → exportDocument round trip in headless Chromium, with the smoke not skipped | `cargo run -p xtask -- dist --web` (bridge smoke stage) | the smoke runs and is not skipped. `dist --web` calls `smoke_bridge`, the release workflow installs headless Chrome and fails closed without it, and `FIRECRAFT_DIST_SKIP_SMOKE` appears nowhere in `.gitforge.yml` or the GitHub workflows — only as a comment at `xtask/src/dist.rs:52`. The rehearsal's `build artifacts` job (run 44188107435) built the bundle and ran the smoke against it: the job log reads `[xtask] dist web: 7 file(s), wasm 15780814 B …, smoke green` and `[xtask] dist --web ok` at 2026-11-17T22:31:41Z. The GitForge `release` job logs the same line for the mirror builds. Receipt: [RECEIPT-0.2.5-DRYRUN-6.md](../release/RECEIPT-0.2.5-DRYRUN-6.md). | — | 2026-11-17: the smoke ran green against the rehearsal's bundle (run 44188107435) | +| T27-50 | Reconcile the production Pages deployment against the release bundle | P0 | SMOKE TESTED | T27-48 | The Pages project, deployment id, deployed asset inventory and per-file hashes are compared to the tagged bundle, and the result is written as a machine-readable report | `curl -s https://firecraft-web.pages.dev/MANIFEST.sha256` | the comparison is a tracked tool and a machine-readable report. `scripts/reconcile_deployment.py` fetches the deployment's `MANIFEST.sha256` and every file it names, hashes the bytes it actually received, and compares them file by file with the bundle's manifest; `scripts/reconcile_deployment-selftest.py` covers the four ways the two can disagree. The report against the live deployment is [`docs/release/DEPLOYMENT-RECONCILIATION.json`](https://github.com/aliasfoxkde/FireCraft/blob/main/docs/release/DEPLOYMENT-RECONCILIATION.json) and its result is `differences`: the live deployment serves bundle `2274ac8c2925b61d` while the v0.2.5 release asset carries `f9dbe2f14305b2d9`, each asset name appearing only on its own side of the inventory; `appinfo.json` matches, and `index.html` and `host.html` differ only by the substituted asset names, their integrity digests and the Pages analytics beacon. **Re-verified 2026-11-17**: the deployment's manifest digest is unchanged at `61e3aeda…` and the comparison against the rehearsal's web bundle reports eight differences — the live bundle has not moved and still corresponds to no identifiable release. | no `CLOUDFLARE_API_TOKEN`; the deploy step in `release.yml` is commented out — a deployment that names a commit stays maintainer-gated (B-11, B-12) | 2026-11-17: the comparison re-run; the live deployment is unchanged and unreconciled | +| T27-51 | Build and verify the standalone CLI archive | P0 | SMOKE TESTED | T27-06 | The CLI archive builds from the candidate commit; architecture, `--version`/`--help`, an import→export round trip, clean extraction and the checksum all verify | `firecraft-cli --version && firecraft-cli import … && firecraft-cli export …` | the rehearsal published `firecraft-cli-0.2.5-linux-x86_64.tar.gz` (8,158,124 B, sha256 `953633e4…`) as a draft asset (run 44188107435). The archive was downloaded from the draft, re-hashed locally (the digest matches the API's) and extracted, and **the extracted binary attests its own build**: `firecraft-cli --version` prints `firecraft-cli 0.2.5` and exits 0. The distro matrix runs the CLI on all six cells ([LINUX-COMPAT.md](../release/LINUX-COMPAT.md)). The archive is byte-identical across two independent rehearsals. | — | 2026-11-17: the published CLI archive attests its own version | +| T27-52 | Generate the source archive from the exact tagged commit and clean-build it | P0 | SMOKE TESTED | T27-06 | The archive is generated from the tag SHA, contains the sources, licenses and build inputs, excludes build outputs and secrets, and a clean extraction runs `cargo metadata --locked` and `cargo build --locked` | `cargo run -p xtask -- dist --source` | the generator is implemented and wired into `release.yml`, and the defect the rehearsal-3 archive exposed — a lockfile whose dependency references did not name the stamped version — was fixed by the `xtask stamp` verb and the `xtask stamps` drift gate. The rehearsal published `firecraft-0.2.5-source.tar.gz` (3,176,024 B, sha256 `0cbee184…`); the workflow's own step reported `clean-extract + cargo metadata --locked verified`, and the archive was then downloaded from the draft and extracted on a **second host**: `cargo metadata --locked` succeeds and `cargo build --locked` finished the dev profile in 53.78s on cargo 1.99.1. The archive's only `.claude/` entry is `.allow_config_edits` — the scratch helper that rehearsal 5 published (`c396b18a`) is gone. The source archive is generated from the tagged commit, carries the sources, the licences and the build inputs, and excludes build outputs and secrets. Receipt: [RECEIPT-0.2.5-DRYRUN-6.md](../release/RECEIPT-0.2.5-DRYRUN-6.md). | — | 2026-11-17: the published source archive clean-builds on a second host | ### RP-06 — Supply-chain metadata, provenance, signing, workflow hardening | ID | Title | Pri | Status | Depends on | Acceptance criteria | Verification | Evidence | Blockers | Done | |---|---|---|---|---|---|---|---|---|---| -| T27-53 | Define canonical asset naming and generate every checksum and the release-wide list | P0 | SMOKE TESTED | T27-06 | Asset naming is canonical; every asset has a SHA-256; one release-wide checksum list covers every published file | `cd dist && sha256sum -c RELEASE_SHA256.txt` | [`release.yml`](https://github.com/aliasfoxkde/FireCraft/blob/main/.github/workflows/release.yml) generates `RELEASE_SHA256.txt` for 7 files (no provenance/signature files). The rehearsal published it (718 B, sha256 `6ae766e7…`); downloaded from the draft, its 7 entries match the API digests of the 7 assets it covers — 7/7. **Naming is not yet canonical**: the source archive still ships as `FireCraft--source.tar.gz` while every other asset is lowercase; the rename to `firecraft--source.tar.gz` is implemented and lands with this row's next rebuild | — | 2026-11-11: checksum list verified against the published digests (rehearsal `v0.2.5-dryrun.3`) | -| T27-54 | Generate and verify a machine-readable release manifest | P0 | BUILD VERIFIED | T27-53 | `release-manifest.json` carries version, tag, source SHA, target triple, format, size, SHA-256, run URL and verification status; missing and corrupted assets are rejected | `jq -r '.version,.artifacts[0].sha256' dist/release-manifest.json` | implemented on `feat/release-manifest` (PR #106, merged in the train as `658f015`) per [decision 0089](https://github.com/aliasfoxkde/FireCraft/blob/main/docs/decisions/0089-release-manifest-shape.md): `xtask dist --manifest` writes `dist/release-manifest.json` (schema `firecraft.release-manifest/1`) for the eight published assets — canonical name, path, size, SHA-256, format, target triple and verification status — plus the run's version, tag, source revision, run URL and timestamp; `dist --verify-manifest` re-checks it against the files; `release.yml` publishes it as a ninth asset and validates it against the downloads. Unit tests cover the shape, the determinism and every rejection path. **No CI receipt and no published manifest yet** — the row closes when a rehearsal publishes one. | T27-53's canonical naming is still `NOT STARTED`; the manifest adopts the names the release already enforces, and its asset table moves with T27-53 (decision 0089) | 2026-11-11: implementation merged (PR #106 → #107) | -| T27-55 | Regenerate and validate the CycloneDX SBOM | P0 | BUILD VERIFIED | T27-06 | The SBOM regenerates from the locked graph, validates against the CycloneDX schema, and is compared component-by-component against `Cargo.lock`; it is published and the download is verified | `cargo run -p xtask -- dist --sbom && jq .metadata.component dist/sbom/firecraft-0.2.5-sbom.cdx.json` | the CycloneDX 1.5 schema and the SPDX sibling it points at are vendored under `xtask/schema/`, and `xtask dist --sbom` validates the document it builds against them *before* writing it and compares its component inventory with `Cargo.lock`'s `[[package]]` sections, naming both sides of any difference (decision [0119](../decisions/0119-the-sbom-is-validated-against-the-schema-it-claims.md)). The validator refuses a schema that uses a keyword it does not implement, so a partial implementation cannot report a bad document as valid. The generator's licence field was wrong for every crate whose licence is an SPDX expression — `MIT OR Apache-2.0` is an expression, not an identifier — and the identifier list is now read from the schema itself. The comparison then failed on its first real run — the CI release job reported `in Cargo.lock but not in the SBOM: glob 0.3.4 / libwebp-sys 0.14.4` — because the inventory came from `cargo metadata`'s *resolved* graph (560 packages) while the lockfile *pins* 562: Cargo resolves the lockfile independently of any one feature selection, so optional and platform-specific crates that nothing here enables are pinned but not resolved. The generator now takes the lockfile as the inventory, lists the two pinned-but-unlinked packages as components with an empty dependency edge (pinned, not linked) and the archive digest the lockfile records, and the comparison is a real assertion rather than a tautology against a resolved graph. **Outstanding**: publishing the newly validated SBOM and verifying the download belongs to the next release rehearsal | — | 2026-11-16: schema validation and the `Cargo.lock` comparison implemented and verified, including the lockfile-inventory fix the first CI run exposed | +| T27-53 | Define canonical asset naming and generate every checksum and the release-wide list | P0 | SMOKE TESTED | T27-06 | Asset naming is canonical; every asset has a SHA-256; one release-wide checksum list covers every published file | `cd dist && sha256sum -c RELEASE_SHA256.txt` | [`release.yml`](https://github.com/aliasfoxkde/FireCraft/blob/main/.github/workflows/release.yml) generates `RELEASE_SHA256.txt` for 7 files and names each file explicitly, because the globs over-match. The naming defect the rehearsal-3 receipt found is closed: every published asset now carries the lowercase canonical name (`firecraft-0.2.5-source.tar.gz`, not `FireCraft-…`). The rehearsal published the list (664 B, sha256 `af054b99…`); downloaded from the draft, its 7 entries match the API digests of the 7 assets it covers — 7/7. The list does not cover itself and does not cover the manifest, which is why it holds seven entries for nine assets. The native `MANIFEST.sha256` is byte-identical across two rehearsals. | — | 2026-11-17: the checksum list verified 7/7 against the rehearsal's published digests | +| T27-54 | Generate and verify a machine-readable release manifest | P0 | SMOKE TESTED | T27-53 | `release-manifest.json` carries version, tag, source SHA, target triple, format, size, SHA-256, run URL and verification status; missing and corrupted assets are rejected | `jq -r '.version,.artifacts[0].sha256' dist/release-manifest.json` | `xtask dist --manifest` writes `dist/release-manifest.json` (schema `firecraft.release-manifest/1`) with one entry per published artifact — canonical name, path, size, SHA-256, format, target triple and verification status — plus the run's version, **tag**, source revision, **run URL** and timestamp; `dist --verify-manifest` re-checks it against the files. The rehearsal published the manifest (3,295 B, sha256 `54b8a231…`) and the workflow's aggregate gate validated it against the downloaded assets. **The manifest names its tag and its run**: `tag` `v0.2.5-dryrun.6`, `source_sha` `c396b18afba0a…`, `run_url` `https://github.com/aliasfoxkde/FireCraft/actions/runs/44188107435`, `run_id` `44188107435`, `built_at` `2026-11-17T22:31:44Z`, 8 artifact entries. Both gaps that made that impossible were closed first — the caller now states the tag and `xtask` fails closed when the stated tag does not name the workspace version, and the run URL is reconstructed from `GITHUB_SERVER_URL`, `GITHUB_REPOSITORY` and `GITHUB_RUN_ID` because GitHub Actions exports no `GITHUB_RUN_URL`. | — | 2026-11-17: a published manifest that names its tag and its run | +| T27-55 | Regenerate and validate the CycloneDX SBOM | P0 | SMOKE TESTED | T27-06 | The SBOM regenerates from the locked graph, validates against the CycloneDX schema, and is compared component-by-component against `Cargo.lock`; it is published and the download is verified | `cargo run -p xtask -- dist --sbom && jq .metadata.component dist/sbom/firecraft-0.2.5-sbom.cdx.json` | the CycloneDX 1.5 schema and the SPDX sibling it points at are vendored under `xtask/schema/`, and `xtask dist --sbom` validates the document it builds against them *before* writing it and compares its component inventory with `Cargo.lock`'s `[[package]]` sections, naming both sides of any difference (decision [0119](../decisions/0119-the-sbom-is-validated-against-the-schema-it-claims.md)). The inventory is the lockfile's own, not a resolved graph, so the two pinned-but-unlinked packages are listed as components with an empty dependency edge. The rehearsal published the SBOM (369,745 B, sha256 `e50516b2…`): the workflow's step reported `562 components, 562 locked package(s), schema-valid`, and the aggregate gate validated the download. Two independent rehearsals' SBOMs are identical in every field except `metadata.timestamp` — same components, same dependency graph, same serial number. | — | 2026-11-17: the validated SBOM published and verified | | T27-56 | Generate and independently verify build provenance | P1 | NOT STARTED | T27-53 | A provenance format is defined; each attestation binds an artifact digest to the source SHA, workflow and runner; attestations verify independently | `gh attestation verify --repo aliasfoxkde/FireCraft` | not implemented — provenance is listed as "planned" in the artifact matrix, and no signing or attestation identity is configured. | T27-53; an attestation identity must exist | — | -| T27-57 | Harden the release workflows | P1 | BUILD VERIFIED | T27-18 | Third-party actions are pinned to full commit SHAs; permissions are least-privilege; build, sign and publish are separate; all artifacts bind to one tag SHA; the gate fails on missing outputs, version mismatch, checksum mismatch, unexpected extra artifacts and a skipped required platform; a safe test-release mode exists; the toolchain is pinned to the compiler the gate's floors were measured with | `python3 scripts/validate_workflow_pins.py .github/workflows`; `bash scripts/validate_workflow_pins-selftest.sh`; `bash scripts/classify_release_mode-selftest.sh`; `bash scripts/validate_release_aggregate-selftest.sh`; `bash tests/release-assets-selftest.sh` | **Pinning implemented 2026-11-11** (decision [0099](../decisions/0099-workflow-actions-are-pinned.md)): all 19 action references in `ci.yml`, `release.yml` and `deploy-site.yml` name a 40-hex commit SHA with the version in a trailing comment; `scripts/validate_workflow_pins.py` fails the CI on an unpinned reference, and `scripts/validate_workflow_pins-selftest.sh` demonstrates each failure mode with one deliberately broken input. Least-privilege permissions were already declared (`contents: read` at workflow level, `contents: write` only for the draft-release job) and the aggregate gate already fails closed on a missing, empty, misnamed or digest-mismatched artifact. The deploy trigger no longer includes the workflow's own file, so a workflow edit is not a production publication. **Both remaining criteria implemented 2026-11-12** (decision [0100](../decisions/0100-release-mode-classification-and-a-tested-release-gate.md)): the aggregate gate moved out of inline YAML into `scripts/validate_release_aggregate.sh`, which now also rejects an artifact the release does not publish and requires every required platform (`x86_64-unknown-linux-gnu`, `wasm32-unknown-unknown`) to be represented in the release manifest; `scripts/validate_release_aggregate-selftest.sh` breaks a complete artifact set one thing at a time — missing, empty, extra artifact, checksum mismatch, manifest version mismatch, manifest digest mismatch, an unlisted asset, an unknown schema, a skipped platform, and a platform the manifest does not represent. A safe test-release mode exists: `scripts/classify_release_mode.sh` maps `vX.Y.Z` to production and `vX.Y.Z-dryrun.N` to a rehearsal and refuses every other shape; a manual `workflow_dispatch` run may only rehearse; a rehearsal release is a prerelease titled `TEST ` whose notes carry a do-not-publish banner, and its disposable tag is created at the commit that built the artifacts. `scripts/classify_release_mode-selftest.sh` demonstrates each refused shape, and `tests/release-assets-selftest.sh` drives the `gh` argument vector against a stub. **Toolchain pinned 2026-11-12** (decision [0102](../decisions/0102-coverage-floor-is-measured-with-a-pinned-toolchain.md)): the workflows install `toolchain: 1.99.1` and the CI image is `rust:1.99.1-slim-bookworm`, because the unpinned `stable` moved the coverage floor silently — rustc 1.100.0 measures a byte-identical Rust tree at 77.0%/76.1% where 1.99.1 measures 81.9%/78.9%, while the GitForge image stayed on 1.99. The gate now names the compiler that measured it. **Outstanding**: the rehearsal path has not been run end to end through `workflow_dispatch` (it needs this workflow on `main`); the manual rehearsal in issue #99 predates the `TEST` title and the prerelease flag | — | 2026-11-12: `gate` green on head `002eb56` (PR #130, merged 2026-11-12) | +| T27-57 | Harden the release workflows | P1 | SMOKE TESTED | T27-18 | Third-party actions are pinned to full commit SHAs; permissions are least-privilege; build, sign and publish are separate; all artifacts bind to one tag SHA; the gate fails on missing outputs, version mismatch, checksum mismatch, unexpected extra artifacts and a skipped required platform; a safe test-release mode exists; the toolchain is pinned to the compiler the gate's floors were measured with | `python3 scripts/validate_workflow_pins.py .github/workflows`; `bash scripts/validate_workflow_pins-selftest.sh`; `bash scripts/classify_release_mode-selftest.sh`; `bash scripts/validate_release_aggregate-selftest.sh`; `bash tests/release-assets-selftest.sh` | the pinning, the least-privilege permissions, the fail-closed aggregate gate, the release-mode classifier and the pinned toolchain are all in place and tested (decisions [0099](../decisions/0099-workflow-actions-are-pinned.md), [0100](../decisions/0100-release-mode-classification-and-a-tested-release-gate.md), [0102](../decisions/0102-coverage-floor-is-measured-with-a-pinned-toolchain.md)). **The remaining criterion is met**: the rehearsal path has now been run end to end through `workflow_dispatch` — twice, on runs 44185273916 (`v0.2.5-dryrun.5`) and 44188107435 (`v0.2.5-dryrun.6`), both `completed`/`success`, each producing a `TEST`-titled prerelease draft with the do-not-publish banner, each classified as a rehearsal by `scripts/classify_release_mode.sh`, each passing the aggregate gate. A manual run may only rehearse, and the rehearsal's artifacts are bound to one tag SHA. | — | 2026-11-17: the rehearsal path ran end to end through `workflow_dispatch` | ### RP-07 — Release aggregation and publication @@ -252,14 +252,14 @@ condition/verification. **RP** is the release-plan work package the row belongs |---|---|---|---|---|---|---|---|---|---| | T27-17 | Replace the invalid `gh release create` asset arguments | P0 | BUILD VERIFIED | T27-03 | Assets are passed as positional filenames; the invocation is tested for missing, duplicate and space-containing filenames | `gh release create --help` | [`release.yml`](https://github.com/aliasfoxkde/FireCraft/blob/main/.github/workflows/release.yml) used `assets+=(-a "$f")`; `gh release create` has no `-a` flag in gh 2.46 — assets are positional `[...]`. **Fixed and exercised**: the rehearsal's `create draft release` job (run 42937560976) succeeded and the draft carries 8 uploaded assets, so the positional invocation works against the real CLI. | — | 2026-11-11: rehearsal `v0.2.5-dryrun.3`, run 42937560976 | | T27-18 | Scope release permissions and verify the draft-to-published transition | P0 | IN PROGRESS | T27-17 | `contents: write` exists only on the publishing job; a draft is created, assets are uploaded, post-upload digests are validated, and the draft transitions to published | `grep -n 'permissions:' -A3 .github/workflows/release.yml` | `release.yml` build job `contents: read`, release job `contents: write` — compliant. **Half exercised**: the rehearsal created a draft with 8 uploaded assets and validated the digests, and the draft is still `draft: true` / `published_at: null`. The draft-to-published transition is untested because publishing is a maintainer-only act (campaign §20). | maintainer approval for a published release | 2026-11-11: draft creation and digest validation verified (run 42937560976); the published transition awaits authorization | -| T27-19 | Exercise the whole draft-release flow on a disposable test tag | P0 | SMOKE TESTED | T27-17, T27-18 | A disposable (non-production) tag produces a draft release whose assets download and match `RELEASE_SHA256.txt` | `gh release download --dir artifacts && (cd artifacts && sha256sum -c RELEASE_SHA256.txt)` | the disposable tag `v0.2.5-dryrun.3` → annotated tag `9009d404…` → commit `384687f…` ("Disposable rehearsal tag (not a release)") produced draft release 441136236 with 8 assets, all `state: uploaded`. The downloaded `RELEASE_SHA256.txt` matches the API digests of all 7 assets it covers. The source archive was downloaded and extracted too — and it does **not** clean-build (T27-52). Receipt: [RECEIPT-0.2.5-DRYRUN-3.md](../release/RECEIPT-0.2.5-DRYRUN-3.md). The row's own removal step — deleting the draft release and the three disposable `v0.2.5-dryrun.*` tags — has been carried out: on 2026-11-16 the releases API answered 404 for the draft release 441136236 and for the tag `v0.2.5-dryrun`, `gh release list` listed no dryrun release, and `git ls-remote --tags origin` listed no dryrun tag. The `test/disposable-release-dryrun` branch is gone from `origin` as well, so the rehearsal commits are no longer reachable from the remote, and issue #99's receipt — the run id, the asset list and its digests — is what remains of the rehearsal. Issue #99 recorded the removal as deferred pending maintainer authorization and carries no authorization note, so this row records the remote's verified state rather than an authorized step. | — | 2026-11-11: rehearsal `v0.2.5-dryrun.3`, run 42937560976 | +| T27-19 | Exercise the whole draft-release flow on a disposable test tag | P0 | SMOKE TESTED | T27-17, T27-18 | A disposable (non-production) tag produces a draft release whose assets download and match `RELEASE_SHA256.txt` | `gh release download --dir artifacts && (cd artifacts && sha256sum -c RELEASE_SHA256.txt)` | the disposable tag `v0.2.5-dryrun.6` → commit `c396b18afba0a…` produced draft release 449279892 with 9 assets, all `state: uploaded`; the draft is `draft: true`, `prerelease: true`, titled `TEST 0.2.5-dryrun.6`, with `published_at: null`. The downloaded `RELEASE_SHA256.txt` matches the API digests of all 7 assets it covers, and the source archive, the checksum list and the manifest were each re-hashed locally — 3 of 3 match. The tag is **lightweight** (`GET /git/ref/tags/…` reports an object of type `commit` and `GET /git/tags/…` answers 404), so it carries no message of its own; the rehearsal's do-not-publish intent lives in the draft state, the `TEST` title and the notes banner. Receipt: [RECEIPT-0.2.5-DRYRUN-6.md](../release/RECEIPT-0.2.5-DRYRUN-6.md). The row's own removal step was carried out on 2026-11-16 for the rehearsal-3 artifacts (the releases API answered 404 for draft 441136236 and for the tag `v0.2.5-dryrun`); **the three later rehearsals are still on the remote** — drafts 449261624 (`v0.2.5-dryrun.5`) and 449279892 (`v0.2.5-dryrun.6`), the rehearsal-4 draft, and their three disposable tags. | the removal of the rehearsal-4, -5 and -6 drafts and tags is outstanding and stays maintainer-gated, as the rehearsal-3 removal was | 2026-11-17: rehearsal `v0.2.5-dryrun.6`, run 44188107435, draft 449279892 | | T27-58 | Make the aggregator fail closed on every invalid artifact shape | P0 | BUILD VERIFIED | T27-17 | The aggregator rejects missing, mislabeled, empty, duplicate, wrong-version, wrong-commit and checksum-mismatched artifacts, and fails when a required platform job was skipped | `bash scripts/validate_release_aggregate-selftest.sh` | `8fc3694` added fail-closed aggregate validation; rehearsed locally negative and positive. **The positive path now runs for real**: the rehearsal's `create draft release` job validated all nine required artifacts, ran `sha256sum -c RELEASE_SHA256.txt`, checked the release manifest against the downloaded assets and every asset name against the tag version, then created the draft (run 42937560976). **2026-11-12**: the aggregator moved into [`scripts/validate_release_aggregate.sh`](https://github.com/aliasfoxkde/FireCraft/blob/main/scripts/validate_release_aggregate.sh) and is now exercised by `scripts/validate_release_aggregate-selftest.sh`, which breaks a complete artifact set one thing at a time — a missing artifact, an empty artifact, an unexpected extra artifact, a checksum mismatch, a manifest version mismatch, a manifest digest mismatch, an asset the manifest does not list, an unknown manifest schema, a skipped required platform, and a platform the manifest does not represent. The last is the discriminating case for the platform check: every artifact is present and every digest agrees, and the gate still fails | — | 2026-11-12: every failure mode demonstrated with one deliberately broken input | ### RP-08 — Final certification | ID | Title | Pri | Status | Depends on | Acceptance criteria | Verification | Evidence | Blockers | Done | |---|---|---|---|---|---|---|---|---|---| -| T27-60 | Produce the release certification report for the approved scope | P0 | NOT STARTED | T27-19, T27-50 | The report names the tag, source SHA, workspace version, approved matrix, every artifact with size/SHA-256/runner/toolchain, install/launch/smoke evidence, signature/SBOM/provenance status, deployment reconciliation, CI links, limitations, unsupported targets, residual risk and independent verification results | `test -f docs/release/CERTIFICATION.md` | no certification report exists — the rehearsal receipt `docs/release/RECEIPT-0.2.5-DRYRUN-3.md` certifies nothing: no artifact is signed and no provenance or attestation exists, so the ladder's ceiling for this repository is rung 8. | — | — | +| T27-60 | Produce the release certification report for the approved scope | P0 | SMOKE TESTED | T27-19, T27-50 | The report names the tag, source SHA, workspace version, approved matrix, every artifact with size/SHA-256/runner/toolchain, install/launch/smoke evidence, signature/SBOM/provenance status, deployment reconciliation, CI links, limitations, unsupported targets, residual risk and independent verification results | `test -f docs/release/CERTIFICATION.md` | the report exists: [[CERTIFICATION.md](../release/CERTIFICATION.md)](../release/CERTIFICATION.md). It names the tag, the source SHA, the workspace version, the approved matrix, every artifact with its size, digest, runner and toolchain, the install/launch/smoke evidence, the signature/SBOM/provenance status, the deployment reconciliation, the CI gates, the limitations, the unsupported targets, the residual risk and the independent verification results. The independent checks are real: `RELEASE_SHA256.txt` downloaded from the draft matches 7 of 7 API digests; three downloaded assets re-hash 3 of 3; the source archive clean-builds on a second host; the native artifacts are byte-identical across two independent rehearsals. **The report certifies nothing beyond that**: nothing is signed and no provenance exists, so no artifact reaches rung 9 of the evidence ladder. | B-14 — no signing identity and no provenance generator exist, so no artifact can be `CERTIFIED`; and approved-scope item 8 (the verified production deployment) is blocked by B-11/B-12 | 2026-11-17: the certification report published in the repository | | T27-61 | Obtain publication authorization | P0 | BLOCKED | T27-60 | The maintainer authorizes publication in writing, and only then is the draft published | `gh release edit --draft=false` | Campaign §20 forbids publishing a production release without explicit maintainer approval | Maintainer authorization | — | ### Remaining rows (Phase-2 PR #8 remediation) @@ -267,7 +267,7 @@ condition/verification. **RP** is the release-plan work package the row belongs | ID | Title | Pri | Status | Depends on | Acceptance criteria | Verification | Evidence | Blockers | Done | |---|---|---|---|---|---|---|---|---|---| | T27-24 | Verify the remaining PR #8 findings, fix the valid ones and document the rejected ones | P1 | IMPLEMENTED | T27-03 | Every remaining Major/Critical finding is independently verified against current code; valid ones are fixed; rejected ones carry a reason | `cargo test -p app -p engine` | All 9 code findings and 25 documentation findings in PR #8's inline comments were adjudicated against `main` `f1a45bec`. Every code finding is already correct on `main`, so each is rejected with the code that is right: `editor.rs:619` `drag_quadrangle` moves the grabbed corner and the next adjacent corner (`let adjacent = (grabbed + 1) % 4;`), which preserves `TL + BR = TR + BL` and matches its doc comment; `editor.rs:2376` Scale, Skew and Distort read `self.live_selection()` (`editor.rs:2494`, `:2525`, `:2563`, `:2587`), not `document().object_selection`; `editor.rs:2595` Redraw Path sends the sampled pointer trace (`self.redraw_selected_path(&drag.trace)` at `editor.rs:2614`; the signature takes `trace: &[geom::Point]` at `:2733`), guarded by `redraw_path_uses_the_sampled_pointer_trace`; `panels_web.rs:595` the `symbol.createInstance` payload carries only `symbol`, `x`, `y`, `page`, `layer` and `cell`, so `CreateInstancePayload`'s `deny_unknown_fields` is satisfied; `panels_web.rs:574` the placement is `symbol_insert_translation` (`panels_web.rs:275`), which centres the symbol's `definition_content_bounds` on the canvas, guarded by the test at `panels_web.rs:1522`; `retouch.rs:656` the four retouch commands call `brush_diameter` (`retouch.rs:118`), which `checked_mul(2)` and returns `EngineError::BadPayload`; `retouch.rs:1149` `retouch_commands_reject_overflowing_brush_diameters` asserts `Err(EngineError::BadPayload { .. })` for all four; `xtask/src/main.rs:222` `main` carries no `install_pattern_state`, no `temp_dir()` and no `firecraft-aegis-home` — the Aegis leg runs the scanner over the repo root against the content-keyed `.aegis-baseline.tsv` (decision 0074); and the Vector Path node drag that a pure click never ends is fixed on `main` by `79fe2759` (#168), which opens the drag in the press routing (`canvas_ui.rs:1060` and `:1111` → `begin_node_reshape`) and closes it on `drag_stopped` (`canvas_ui.rs:1193` → `:1205`), so no transaction leaks. The one valid code finding, `editor.rs:3271-3276` draft tool ownership, is fixed on `main` by `d3e45cf9` (PR #171) with three regression tests. Of the 25 documentation findings, 23 were already resolved on `main`: the trademark notice at `README.md:22`, `-p app-cli` at `README.md:102` and `:112`, the live-ledger pointer at `README.md:55`, the historical-gaps label at `README.md:156`, the six-cell `1280×800` row and the reconciled menu-resource evidence in `docs/CS6-PARITY.md`, the panel-status corrections in `docs/NEXT.md`, both optimized builds in `docs/PROGRESS.md`, the `ToolPalette`/`Slices` and `57651` evidence notes in `docs/research/fw-ui-evidence.md`, the remaining-work list in `docs/HANDOFF.md`, the L4 crate row and the right-side-dock description in `docs/wiki/`, the closed menu burn-down in `docs/TASKS.md`, the Aegis baseline re-blessed for decision 0083, and the conflict-free `docs/HANDOFF.md:3` orientation. Two remained valid and are fixed here: `docs/wiki/Parity-Roadmap.md` now states that the 2-inert-of-298 census is not a coverage number and names the 35 `ShellAction::ReportGap` rows, and `README.md:60` labels `docs/GAPS.md` the historical gaps audit instead of current gaps. The per-finding table is in [`docs/tasks/T27-24-pr8-remaining-findings.md`](https://github.com/aliasfoxkde/FireCraft/blob/main/docs/tasks/T27-24-pr8-remaining-findings.md). The same change repairs the Aegis baseline and closes the gap that let it drift: PR #170 merged `docs/tasks/T27-41-windows-arm64-assessment.md` through a documentation-only path, and both CI systems skip the Aegis leg for documentation-only changes — GitHub's `gate` job is gated on `code_changed` while its `docs-only-check` job validates whitespace only, and the GitForge pipeline has no Aegis job at all — so four Windows build literals entered `main` unblessed, and the next code change's gate failed for content it never touched (`main` `c26adffc`: 5 findings the baseline does not cover). The literals are adjudicated as false positives in the `zip-code` class note (`xtask/src/aegis.rs`), the baseline is re-blessed (24 keys added — 18 `zip-code`, 4 `ssrf`, 2 `ssrf-localhost` — each with a reason in the table below), and the scan now runs on documentation-only changes in both CI systems — [decision 0132](../decisions/0132-the-secret-scan-runs-on-documentation-changes.md), row T27-62. | — | 2026-11-17 | -| T27-25 | Run the full application suite, the CI gate and the web smoke for PR #8 | P0 | NOT STARTED | T27-23, T27-24 | The app test suite, `xtask ci`, the web build and the browser smoke all pass on the PR head | `cargo run -p xtask -- ci && cargo run -p xtask -- dist --web` | none yet — PR #8 was closed unmerged; its review findings are re-hosted in issue #47 and the parity gate in closed issue #86. | T27-23, T27-24 | — | +| T27-25 | Run the full application suite, the CI gate and the web smoke for PR #8 | P0 | SMOKE TESTED | T27-23, T27-24 | The app test suite, `xtask ci`, the web build and the browser smoke all pass on the PR head | `cargo run -p xtask -- ci && cargo run -p xtask -- dist --web` | the whole suite ran on the head `f710c524` and passed. PR #8 itself was closed unmerged on 2026-11-12, so the head that carries its re-hosted findings (PR #171, merged as `d3e45cf9`) is the branch head `f710c524`, and that is what was run. **GitHub**, 13 check runs, all terminal on that SHA (the PR run is [44194696167](https://github.com/aliasfoxkde/FireCraft/actions/runs/44194696167)): the two `gate` legs — each `cargo run --package xtask -- ci-no-fmt`, the canonical gate (format, clippy with `-D warnings`, the workspace test suite, layer enforcement, the wasm build) — both `success`; the two `fmt` legs `success`; the two `classify-change` legs `success`; the two `docs-only-check` legs `skipped` because this change is not documentation-only, and the Aegis secret scan runs in its own legs (`cargo run --locked --package xtask -- aegis`), which are `success`; CodeQL `success`; CodeRabbit `success`; the three `Analyze` legs `success`. **GitForge mirror**, pipeline run `65de365f` on the same SHA: all 12 jobs `succeeded` — `clippy-test` (the application test suite), `wasm-web` (`[xtask] wasm ok`), `release` (`[xtask] dist web: 6 file(s), wasm 15783798 B (warning 25000000, hard ceiling 50000000), manifest MANIFEST.sha256, smoke green` and `[xtask] dist --web ok` — the browser smoke against the built bundle), `aegis`, `layers`, `fmt-layers`, `corpus-golden`, `fuzz`, `perf-budgets`, `generated-docs` and `release-ledger` (the committed ledger has no problems). | — | 2026-11-17: the full gate, the application test suite, the web build and the browser smoke all green on the head `f710c524` | | T27-23 | Track the originating tool in plotted-draft state, and cover it with regression tests | P0 | IMPLEMENTED | T27-03 | Draft state records its owning plotting tool; switching tools resets or explicitly converts the draft; regression tests cover Pen→Hotspot and Hotspot→Pen | `cargo test -p app` | CodeRabbit Major on `crates/app/src/editor.rs:3271-3276` (Pen points folded into a Hotspot commit) — reproduced against the current code and fixed: the pending draft now records the tool that plotted it (`PointDraft::tool`), a tool switch converts the pending path explicitly (committed as the tool that plotted it, dropped when it has too few points to become an object), and `commit_draft` dispatches on the draft's own tool rather than the selected one. The regressions are in [`crates/app/tests/plotted_draft.rs`](https://github.com/aliasfoxkde/FireCraft/blob/main/crates/app/tests/plotted_draft.rs) — Pen→Hotspot, Hotspot→Pen and the too-few-points drop. Merged as PR #171 → `d3e45cf9`; the GitHub `gate`, `fmt` and `classify-change` checks are green on that head and the GitForge pipeline for it (`93905d80`, head `7ff1321e`) succeeded. | — | 2026-11-17 | ### PR #8 finding census (all PR #8 inline comments, for T27-23/24/25) diff --git a/docs/release/LINUX-COMPAT.md b/docs/release/LINUX-COMPAT.md new file mode 100644 index 00000000..1b50b85d --- /dev/null +++ b/docs/release/LINUX-COMPAT.md @@ -0,0 +1,204 @@ +# Linux platform compatibility + +> The authoritative statement of what FireCraft is **tested** on, what +> is only **built**, and what is **deferred**. Read it before assuming a +> Linux artifact works on a given machine. +> +> Authority: [decision 0075](../decisions/0075-release-scope-mvp.md) +> (release scope, accepted) and the artifact inventory in +> [ARTIFACT-MATRIX.md](ARTIFACT-MATRIX.md). Execution ledger: +> [EXECUTION-BOARD.md](EXECUTION-BOARD.md) rows T27-33 and T27-34. +> +> A cell is listed as *tested* only when the artifact was installed (or +> unpacked) on that exact distribution **and** the desktop binary was +> launched there. *Built* means the bytes exist but were never exercised +> on that platform. *Deferred* means a recorded decision took it out of +> scope. Nothing is implied: an absent row is an untested combination, +> and every untested combination is listed. + +## What was exercised + +**Date:** 2026-11-17 · **Host:** Fedora Linux 44 (Workstation Edition), +x86_64, Docker 29.7.2 · **Artifacts built from:** `a8a59b35` on the same +host, inside `rust:1.99-slim-bookworm` (Debian 12), so the DEB's glibc +floor is the real one. + +**These are the published bytes.** The three artifacts tested here are +byte-identical to the assets the release workflow published in the +rehearsal `v0.2.5-dryrun.5` (draft release 449261624): the digests +below are the digests the GitHub API reports for those assets. The +matrix is therefore a receipt for the *published* artifacts, not for a +local rebuild that merely resembles them. + +| Artifact | Bytes | sha256 | +|---|---|---| +| `firecraft_0.2.5_amd64.deb` | 16,228,104 | `7928369626375dfca80516551870d0ab7c98b3f3021ff5c82fc36a86dc39fa42` | +| `firecraft-0.2.5-linux-x86_64.tar.gz` | 24,389,908 | `cb55a52507477c9af327c35be8bb0e6d7dfc16a767e34bfde1aab0d92f346a3d` | +| `firecraft-cli-0.2.5-linux-x86_64.tar.gz` | 8,158,124 | `953633e40c2021858e7001f9691a47ad4ed4491ec752b38ccd716f8e99339093` | + +The DEB declares `Version: 0.2.5`, `Architecture: amd64` and +`Depends: libc6 (>= 2.36), libgcc-s1, libgl1, libx11-6, libxcursor1, +libxrandr2, libxi6, libxkbcommon0, libxkbcommon-x11-0`. + +### The tested matrix + +| Distribution | Version | Arch | DEB | tar.gz (desktop) | tar.gz (CLI) | X11 | Wayland | +|---|---|---|---|---|---|---|---| +| Ubuntu | 24.04.5 LTS | x86_64 | **tested** | **tested** | **tested** | **tested** | **tested** | +| Debian | 12 (bookworm) | x86_64 | **tested** | **tested** | **tested** | **tested** | **tested** | +| Fedora | 45 | x86_64 | n/a — format | **tested** | **tested** | **tested** | **tested** | + +"n/a — format" is a fact about the artifact, not a gap in the test: a +`.deb` is a Debian package format, and no package manager on a +Fedora-family system can install it. The Fedora cells therefore test +the format-agnostic artifact, the desktop tar.gz. + +### Formats that have no tested row + +| Format | State | Why | +|---|---|---| +| AppImage | **never built** | the recipe exists (`packaging/build-appimage.sh`) but has never been executed — B-8 (`linuxdeploy` absent, no pinned copy vendored), B-10 (the recipe ignores `CARGO_TARGET_DIR`), B-9 (the icon is in-tree but `linuxdeploy` has still never run) | +| Flatpak | **never built** | the recipe and manifest exist (`packaging/build-flatpak.sh`, `packaging/flatpak/dev.firecraft.FireCraft.yml`) but have never been executed — B-11 (`flatpak-builder` absent), plus the manifest's module source is the local working tree, so a bundle built from it would not be reproducible from a tag SHA | +| RPM | **deferred** | [decision 0070](../decisions/0070-rpm-deferred.md), re-validated 2026-11-16 (T27-31): no issue on the tracker asks for an RPM; Fedora-family users are served by the tar.gz | + +## What each cell actually did + +The matrix is reproducible: `scripts/release-matrix.sh ` +on any Docker host, with the leg scripts in `packaging/linux/`. + +**Debian-family cells** (Ubuntu 24.04, Debian 12) — install the DEB with +`apt`, so apt itself resolves the package's declared dependencies, then +run the repository's own eight-stage package verification +(`packaging/deb/verify-install.sh`): + +1. metadata — `Version: 0.2.5`, `Architecture: amd64` +2. desktop entry — installed, names `firecraft`/`firecraft`, and + `desktop-file-validate` accepts it +3. icon — installed at `/usr/share/icons/hicolor/256x256/apps/firecraft.png` +4. CLI — `firecraft-cli` runs +5. CLI document round trip — `firecraft-cli info` reads the document, + `firecraft-cli export` writes a 97 B PNG with the signature + `89504e470d0a1a0a` +6. GUI launch under Xvfb — the process stayed up for the 25 s launch + window (exit status 124) +7. upgrade — **not exercised**: no previous package was supplied +8. uninstall / reinstall — removal took the binaries, the desktop entry + and the icon with it; reinstall restored the payload and the CLI runs + +Then the desktop tar.gz is unpacked and launched under Xvfb (exit 124), +and `fc-list` enumerates fonts. + +**Fedora-family cells** (Fedora 45) — no `dpkg`, so the DEB is recorded +as not applicable; the desktop tar.gz is unpacked, `firecraft-cli +--help` runs, the desktop binary launches under `xvfb-run` (exit 124), +and `fc-list` enumerates fonts. + +**Wayland cells** (all three families) — weston's headless backend +serves `wayland-0`, the desktop binary is launched with +`WAYLAND_DISPLAY=wayland-0` and stays up for the 25 s launch window +(exit 124), and `fc-list` enumerates fonts. + +All six cells passed on 2026-11-17. The cell logs are the receipts. + +The matrix was run twice on that date. The first run exercised copies of +the leg scripts; the second run — the one the receipts above describe — +used the repository's own `scripts/release-matrix.sh` and +`packaging/linux/leg-*.sh` at commit `8f9a602c`, with the same six +cells and the same artifacts. The second run exists because the first +run's copies were not the tracked scripts, and because the tracked +Wayland leg had a real defect the copies did not have: it ran +`apt-get install` on a fresh image whose package lists were never +fetched, so the Debian and Ubuntu Wayland cells answered "Unable to +locate package weston" and failed. The leg now refreshes the lists +first, and all six cells pass with the tracked scripts. + +## What is NOT tested + +These are gaps, not endorsements. Nothing in this table claims support +for a combination that is not in the tested matrix above. + +**Distributions and versions** — every other release of every family: +Ubuntu 22.04 LTS, 20.04 LTS and every earlier release; Debian 13, 11, +10 and every earlier release; Fedora 44 down to 40; RHEL / CentOS +Stream / AlmaLinux / Rocky; Arch and derivatives; openSUSE / SLES; +Alpine; NixOS; Clear Linux; Gentoo; Void; Solus; any non-Debian, +non-Fedora family. + +**Architectures** — every architecture other than x86_64: aarch64/arm64 +(assessed feasible-later, [T27-32](../tasks/T27-32-linux-aarch64-assessment.md), +deferred), 32-bit x86, RISC-V, ppc64le, s390x. + +**Real hardware** — every cell ran in a container with software +rendering (llvmpipe/swrast). No cell used a real GPU, a vendor +OpenGL/Vulkan driver, a Wayland compositor with hardware acceleration, +or a display manager. + +**Real desktop sessions** — X11 was Xvfb (a virtual framebuffer) and +Wayland was weston's headless backend. No cell ran GNOME, KDE Plasma, +Xfce, Cinnamon or a systemd user session. Desktop integration was +checked at the *package* level only: the entry is installed and +validates, and the icon is installed. It was **not** checked at the +*session* level: no application menu appearance, no file association, +no thumbnailer, no MIME registration, no icon-theme resolution beyond +the file's presence. + +**Interaction and I/O** — no cell exercised audio, clipboard, drag and +drop, file dialogs, HiDPI scaling, multi-monitor layouts, GPU +acceleration paths, or the Vulkan backend. + +**The DEB's upgrade path** — stage 6 of the package verification needs a +previous package to upgrade from, and none was supplied, so the upgrade +transition is untested. + +**The published v0.2.5 release assets** — the artifacts tested here are +built from `a8a59b35` in this repository, and they are byte-identical +to the assets the rehearsal published. The *production* `v0.2.5` +release's own assets were published out of band (B-21): they were +downloaded and checksum-verified on 2026-11-09, but their clean +container receipts were taken against the **v0.2.3** package, not +against those bytes, so the production DEB still carries the open +defects below. + +## Known defects on these platforms + +- **B-7 — the published DEB's glibc floor.** The DEB tested here + declares `libc6 (>= 2.36)` and the build gate + (`xtask dist --native`) fails the build when the declared floor is + below the binaries' real maximum GLIBC symbol version, read back with + `objdump`. The **published** v0.2.5 package under-declares its floor; + that is an open defect on the published asset, not on this build. +- **B-24 — the published DEB has no desktop integration.** The published + v0.2.5 package carries only the two binaries and the copyright file: + no desktop entry, no icon. The DEB tested here has both (stages 2 and + 3 pass), but the published asset still does not. +- **Issue #39 — the published DEB declares only `libc6`.** The DEB + tested here declares the full X11/GL runtime set + (`libgl1`, `libx11-6`, `libxcursor1`, `libxrandr2`, `libxi6`, + `libxkbcommon0`, `libxkbcommon-x11-0`), so the Fedora-family cells + install their own runtime stack. The published package does not, so + it installs cleanly and then fails to start on a system without that + stack already present. + +## Reproducing this + +```sh +cargo run -p xtask -- dist --native # produces dist/native/ +scripts/release-matrix.sh dist/native # six cells, one per row +``` + +The matrix needs a Docker host. On a SELinux-enforcing host the bind +mounts carry `:z`; without it the containers cannot read the mount and +every cell fails with "Permission denied" before it runs anything. + +## Relationship to the other release documents + +- [ARTIFACT-MATRIX.md](ARTIFACT-MATRIX.md) — one row per artifact, with + its evidence state and blockers. +- [EXECUTION-BOARD.md](EXECUTION-BOARD.md) — the per-unit ledger; + T27-33 is the matrix, T27-34 is this document. +- [RECEIPT-0.2.5-DRYRUN-6.md](RECEIPT-0.2.5-DRYRUN-6.md) — the + rehearsal receipt for the release workflow path, and the run that + published the artifacts this table tests. +- [RECEIPT-0.2.5-DRYRUN-3.md](RECEIPT-0.2.5-DRYRUN-3.md) — the first + rehearsal receipt; it found the source-archive and naming defects + this table's artifacts no longer carry. diff --git a/docs/release/RECEIPT-0.2.5-DRYRUN-6.md b/docs/release/RECEIPT-0.2.5-DRYRUN-6.md new file mode 100644 index 00000000..9fc2af8e --- /dev/null +++ b/docs/release/RECEIPT-0.2.5-DRYRUN-6.md @@ -0,0 +1,233 @@ +# Receipt — the `v0.2.5-dryrun.6` disposable release rehearsal + +> The sixth rehearsal of the tag-driven release workflow, and the one +> whose **published bytes** the Linux compatibility matrix tests. It is +> a **rehearsal**: a disposable tag, a **draft** release, no +> publication. The production release `v0.2.5` is untouched and +> remains the latest published release. +> +> Every fact below was read from the authoritative store at the time of +> writing — the GitHub API for the tag, the workflow run, the draft +> release and its assets, and the workflow's own job logs — not from a +> watcher or from a remembered log. +> +> The earlier receipts are +> [RECEIPT-0.2.5-DRYRUN-3.md](RECEIPT-0.2.5-DRYRUN-3.md) (the first +> end-to-end run) and the rehearsals 4 and 5, whose findings are +> carried forward below. + +## What was run + +| Fact | Value | Where it was read | +|---|---|---| +| Tag | `v0.2.5-dryrun.6` | `GET /git/ref/tags/v0.2.5-dryrun.6` | +| Tag object type | `commit` — the tag is **lightweight** | same | +| Annotated tag object | none — `GET /git/tags/v0.2.5-dryrun.6` answers 404 | same | +| Tagged commit | `c396b18afba0a81f9589b7f74bd412aaca53d62d` | `GET /git/ref/tags/…` → `.object.sha` | +| Workflow trigger | `workflow_dispatch` | `GET /actions/runs/44188107435` → `.event` | +| Release workflow run | **44188107435** — `completed`/`success`, created 2026-11-17T22:14:43Z, head `c396b18a…` | `GET /actions/runs/44188107435` | +| `build artifacts` job | success — 2026-11-17T22:14:47Z → 22:31:49Z | `GET /actions/runs/44188107435/jobs` | +| `create draft release` job | success | same | +| Release mode | **rehearsal** | the run's own log line: `the workspace builds version 0.2.5 for tag v0.2.5-dryrun.6` | +| Draft release | id **449279892**, `draft: true`, `prerelease: true`, `published_at: null` | `GET /releases/449279892` | +| Draft title | `TEST 0.2.5-dryrun.6` | same | +| Draft notes | the `TEST RELEASE — not for publication.` banner and the seven-entry checksum list | same | + +The draft was **not** published. Publishing a production release is a +maintainer-only act (campaign §20); a disposable rehearsal tag and a +draft are what this authorization covers. + +## The nine assets, as published on the draft + +Every asset is `state: uploaded`; the digest is the API's own +`sha256:` field for the asset bytes. + +| Asset | Bytes | sha256 | +|---|---|---| +| `firecraft-0.2.5-linux-x86_64.tar.gz` | 24,389,908 | `cb55a52507477c9af327c35be8bb0e6d7dfc16a767e34bfde1aab0d92f346a3d` | +| `firecraft-cli-0.2.5-linux-x86_64.tar.gz` | 8,158,124 | `953633e40c2021858e7001f9691a47ad4ed4491ec752b38ccd716f8e99339093` | +| `firecraft_0.2.5_amd64.deb` | 16,228,104 | `7928369626375dfca80516551870d0ab7c98b3f3021ff5c82fc36a86dc39fa42` | +| `firecraft-web-0.2.5.zip` | 5,853,417 | `898be851f2e1abac03e0e01a58a3d5f593e74dd44acd572f0d47c4cbfef9d3c4` | +| `firecraft-0.2.5-source.tar.gz` | 3,176,024 | `0cbee1846cb24211fe3be899d286dd1917fd2464d2219f754ae9df7744d5f523` | +| `firecraft-0.2.5-sbom.cdx.json` | 369,745 | `e50516b21fcb5254824a262d73649d746b971a27b1d4a4ffcabe554f0b703174` | +| `MANIFEST.sha256` | 398 | `6a8692ed6806fe9074e8f29762bbbbaabd78ace8bba3e1c5bd5702ead4a5a101` | +| `release-manifest.json` | 3,295 | `54b8a231e2d4ff7e155915867f8d085dcfb86f944815032bd67e06ce2277ee57` | +| `RELEASE_SHA256.txt` | 664 | `af054b99c8c3b864de127597f23bf0f133b4d65f6504de70ba8ede11a61e8381` | + +`RELEASE_SHA256.txt` was downloaded from the draft and compared with +the API's digests above: **7 of 7 entries match** — the desktop +tarball, the CLI tarball, the DEB, the web ZIP, the source archive, +the SBOM and `MANIFEST.sha256`. The list does not cover itself and +does not cover the manifest, so it holds seven entries for nine +assets. + +The source archive, `RELEASE_SHA256.txt` and `release-manifest.json` +were each downloaded and re-hashed locally: **3 of 3 match** the API's +digests byte for byte. + +## The manifest names its tag and its run + +`release-manifest.json`, downloaded from the draft: + +| Field | Value | +|---|---| +| `schema` | `firecraft.release-manifest/1` | +| `version` | `0.2.5` | +| `tag` | **`v0.2.5-dryrun.6`** | +| `source_sha` | `c396b18afba0a81f9589b7f74bd412aaca53d62d` | +| `run_url` | `https://github.com/aliasfoxkde/FireCraft/actions/runs/44188107435` | +| `run_id` | `44188107435` | +| `built_at` | `2026-11-17T22:31:44Z` | +| artifacts | 8 entries, each with its name, size, digest and verification status | + +Two things had to be fixed for this to be possible, and both were +fixed before this rehearsal ran: + +- **The tag.** A rehearsal is built from `main` and its tag is created + *after* the build, so neither the git tag nor the CI ref name can + name it — the manifest came out with `"tag": null` while the release + it described had a tag. The caller now *states* the tag + (`RELEASE_TAG=v0.2.5-dryrun.6` in the workflow's environment) and + `xtask` validates the statement fail-closed: the stated tag must + name the workspace version, either exactly (`v0.2.5`) or with a + disposable suffix (`v0.2.5-dryrun.6`). +- **The run URL.** GitHub Actions does **not** export `GITHUB_RUN_URL`, + so the manifest reconstructed it from the three variables it does + export — `GITHUB_SERVER_URL`, `GITHUB_REPOSITORY` and + `GITHUB_RUN_ID`. + +## What the rehearsal proved + +- **The workflow runs end to end through `workflow_dispatch`.** This + was T27-57's outstanding item: until now the rehearsal path had + never been exercised through the manual trigger, which is the only + trigger a rehearsal is allowed to use. +- **The native set builds and installs.** `xtask dist --native` + produced three archives, verified the DEB with `dpkg-deb`, and + parse-checked nine shell recipes; then the workflow's own Debian 12 + container ran the repository's eight-stage package verification and + **all eight stages passed** — install, desktop entry, icon, CLI, + CLI document round trip, GUI launch, **upgrade**, uninstall and + reinstall. The upgrade stage is exercised here, which the separate + distro matrix does not exercise. +- **The web bundle builds and smokes.** `xtask dist --web` produced 7 + files (wasm 15,780,814 B) and ran the bridge smoke in headless + Chromium: the log reads `smoke green` and `dist --web ok`. The smoke + is not skipped — the workflow installs headless Chrome and fails + closed without it. +- **The source archive is clean.** `xtask dist --source` reported + `firecraft-0.2.5-source.tar.gz (0cbee184…), clean-extract + cargo + metadata --locked verified`. +- **The SBOM is schema-valid and complete against the lockfile.** + `562 components, 562 locked package(s), schema-valid`. +- **The manifest verifies.** `dist --manifest` wrote eight artifact + entries and `dist --verify-manifest` re-checked them against the + files: both steps succeeded. +- **The aggregate gate and the mode classifier both held.** The + `create draft release` job ran `scripts/validate_release_aggregate.sh` + (fail-closed on a missing, empty, extra, misnamed or digest-mismatched + artifact, and on a skipped required platform) and + `scripts/classify_release_mode.sh`, which classified this run as a + rehearsal and produced a `TEST`-titled prerelease with the + do-not-publish banner. + +## Independent verification against the published bytes + +These checks were run against the **downloaded assets**, not against +the build that produced them. + +| Check | Result | +|---|---| +| `RELEASE_SHA256.txt` from the draft vs the API's digests | **7 of 7 match** | +| The three downloaded assets re-hashed locally | **3 of 3 match** the API's digests | +| The source archive extracted, then `cargo metadata --locked --offline` | exit **0** | +| The source archive extracted on a second host (Fedora), then `cargo metadata --locked` and `cargo build --locked` | `METADATA OK`, then `Finished dev profile … in 53.78s` on cargo 1.99.1 | +| The source archive's contents | the only `.claude/` entry is `.allow_config_edits` — **no `scratch/` directory and no `run_engine_test.sh`** | + +## Reproducibility, measured against the previous rehearsal + +`v0.2.5-dryrun.5` (draft 449261624) and `v0.2.5-dryrun.6` (draft +449279892) are two independent runs of the workflow on two commits +that differ only in the release scripts and one removed scratch file. +Their assets compare as follows: + +| Asset | dryrun.5 | dryrun.6 | Same bytes? | +|---|---|---|---| +| desktop tar.gz | `cb55a525…` | `cb55a525…` | **yes** | +| CLI tar.gz | `953633e4…` | `953633e4…` | **yes** | +| DEB | `79283696…` | `79283696…` | **yes** | +| `MANIFEST.sha256` | `6a8692ed…` | `6a8692ed…` | **yes** | +| source archive | `8184bacb…` (3,176,830 B) | `0cbee184…` (3,176,024 B) | no — the commits differ, and dryrun.6 dropped the scratch script (806 B smaller) | +| SBOM | `219684d4…` | `e50516b2…` | no — **only** `metadata.timestamp` differs (`22:11:21Z` vs `22:31:43Z`); the 562 components, the dependency graph and the serial number are identical | +| web ZIP | `5e214be7…` | `898be851…` | no — **the payload is identical**: all seven files inside hash the same as dryrun.5's (`host.html` `379720cc…`, the js `3248368c…`, the wasm `089eada8…`, `index.html` `81b6b805…`, `appinfo.json` `7d039c36…`, `MANIFEST.sha256` `4f47125f…`, `favicon.png` `6e5975dd…`); only the ZIP container's own timestamps differ | +| `release-manifest.json` | `71f78945…` | `54b8a231…` | no — it carries the tag, the run and the build timestamp, which are what it exists to record | + +So the **native artifacts are bit-for-bit reproducible** across +rehearsals, and the two artifacts that are not (the SBOM's timestamp +and the ZIP container's timestamps) differ only in material that is +not content. + +## What the rehearsal found + +**The published source archive no longer ships a scratch script.** +Rehearsal `v0.2.5-dryrun.5` published an archive that contained +`.claude/scratch/run_engine_test.sh` — a session helper that polls the +host's CPU and I/O pressure and then runs one engine test in a +worktree named `/nas/Temp/repos/FireCraft/.claude/worktrees/wt-hist-session`. +It had been force-added before `.claude/scratch/` was ignored, so the +ignore rule did not stop it and `git archive` — which ships exactly +what is tracked — published it as part of the release source. + +The file is untracked now (commit `c396b18a`), and this rehearsal's +archive is the proof: extracted, its only `.claude/` entry is +`.allow_config_edits`. The archive's size dropped by exactly 806 B. + +**The rehearsal tag is lightweight.** `GET /git/ref/tags/v0.2.5-dryrun.6` +reports an object of type `commit`, and `GET /git/tags/v0.2.5-dryrun.6` +answers 404 — there is no annotated tag object and therefore **no tag +message**. Rehearsal `v0.2.5-dryrun.3` could state "the tag says so +in its own message"; that property does not hold for a tag the +workflow creates. The "this is not a release" intent lives in the +draft state, the `TEST` title and the do-not-publish banner — which +is where the rehearsal's safety actually comes from — but the tag +itself is now indistinguishable from a production tag except by its +name. + +**The test target did not compile.** Not a defect of this rehearsal, +but found while it ran: the two manifest tests added for rehearsals 5 +and 6 call `stated_tag_names_version` and `built_run_url` from their +own module without importing them, so `cargo clippy --all-targets` +failed with thirteen `E0425` errors on `c396b18a`. The release build +never compiles the tests, so every rehearsal passed while the test +target was broken. Fixed in `d1029347`. + +## What this receipt does not claim + +- **It does not certify anything.** Nothing in this repository is + signed and no provenance or attestation exists, so no artifact can + reach rung 9 of the evidence ladder. This receipt is evidence, not + a certification; the certification report is + [CERTIFICATION.md](CERTIFICATION.md), and it says the same thing. +- **It does not publish anything.** The release is a draft with + `published_at: null`, and the production `v0.2.5` release is + untouched by this rehearsal. +- **It does not make the live deployment a release.** The live site + still serves a bundle that matches no release; the comparison is in + [DEPLOYMENT-RECONCILIATION.json](DEPLOYMENT-RECONCILIATION.json) and + its result is `differences`. +- **It does not cover the extended scope.** No Windows, macOS, + AppImage, Flatpak, Android, ARM64 or RPM artifact exists. +- **It does not claim the desktop session was tested.** The matrix + that consumes these bytes ran under Xvfb and weston's headless + backend; see [LINUX-COMPAT.md](LINUX-COMPAT.md). + +## Related documents + +- [CERTIFICATION.md](CERTIFICATION.md) — what the approved release + scope can honestly be certified for. +- [LINUX-COMPAT.md](LINUX-COMPAT.md) — the six-cell matrix that + installs and launches these bytes. +- [EXECUTION-BOARD.md](EXECUTION-BOARD.md) — the per-unit ledger. +- [ARTIFACT-MATRIX.md](ARTIFACT-MATRIX.md) — one row per artifact. +- [RECEIPT-0.2.5-DRYRUN-3.md](RECEIPT-0.2.5-DRYRUN-3.md) — the first + rehearsal, which found the source-archive and naming defects. diff --git a/docs/tasks/T27-19-draft-release-test-tag.md b/docs/tasks/T27-19-draft-release-test-tag.md index 65796671..95c3874a 100644 --- a/docs/tasks/T27-19-draft-release-test-tag.md +++ b/docs/tasks/T27-19-draft-release-test-tag.md @@ -19,9 +19,9 @@ whose assets download successfully and every downloaded file matches **Verification:** `gh release download --dir artifacts && (cd artifacts && sha256sum -c RELEASE_SHA256.txt)` -**Evidence:** the disposable tag `v0.2.5-dryrun.3` → annotated tag `9009d404…` → commit `384687f…` ("Disposable rehearsal tag (not a release)") produced draft release 441136236 with 8 assets, all `state: uploaded`. The downloaded `RELEASE_SHA256.txt` matches the API digests of all 7 assets it covers. The source archive was downloaded and extracted too — and it does **not** clean-build (T27-52). Receipt: [RECEIPT-0.2.5-DRYRUN-3.md](../release/RECEIPT-0.2.5-DRYRUN-3.md). The row's own removal step — deleting the draft release and the three disposable `v0.2.5-dryrun.*` tags — has been carried out: on 2026-11-16 the releases API answered 404 for the draft release 441136236 and for the tag `v0.2.5-dryrun`, `gh release list` listed no dryrun release, and `git ls-remote --tags origin` listed no dryrun tag. The `test/disposable-release-dryrun` branch is gone from `origin` as well, so the rehearsal commits are no longer reachable from the remote, and issue #99's receipt — the run id, the asset list and its digests — is what remains of the rehearsal. Issue #99 recorded the removal as deferred pending maintainer authorization and carries no authorization note, so this row records the remote's verified state rather than an authorized step. **Done:** 2026-11-11: rehearsal `v0.2.5-dryrun.3`, run 42937560976. +**Evidence:** the disposable tag `v0.2.5-dryrun.6` → commit `c396b18afba0a…` produced draft release 449279892 with 9 assets, all `state: uploaded`; the draft is `draft: true`, `prerelease: true`, titled `TEST 0.2.5-dryrun.6`, with `published_at: null`. The downloaded `RELEASE_SHA256.txt` matches the API digests of all 7 assets it covers, and the source archive, the checksum list and the manifest were each re-hashed locally — 3 of 3 match. The tag is **lightweight** (`GET /git/ref/tags/…` reports an object of type `commit` and `GET /git/tags/…` answers 404), so it carries no message of its own; the rehearsal's do-not-publish intent lives in the draft state, the `TEST` title and the notes banner. Receipt: [RECEIPT-0.2.5-DRYRUN-6.md](../release/RECEIPT-0.2.5-DRYRUN-6.md). The row's own removal step was carried out on 2026-11-16 for the rehearsal-3 artifacts (the releases API answered 404 for draft 441136236 and for the tag `v0.2.5-dryrun`); **the three later rehearsals are still on the remote** — drafts 449261624 (`v0.2.5-dryrun.5`) and 449279892 (`v0.2.5-dryrun.6`), the rehearsal-4 draft, and their three disposable tags. **Done:** 2026-11-17: rehearsal `v0.2.5-dryrun.6`, run 44188107435, draft 449279892. -**Blockers:** — +**Blockers:** the removal of the rehearsal-4, -5 and -6 drafts and tags is outstanding and stays maintainer-gated, as the rehearsal-3 removal was **Stopping condition:** a successful disposable-tag run is recorded with the tag name, the run id, the downloaded asset list and the checksum verification output. diff --git a/docs/tasks/T27-21-deb-build-and-clean-install.md b/docs/tasks/T27-21-deb-build-and-clean-install.md index a64fd99a..80cbd665 100644 --- a/docs/tasks/T27-21-deb-build-and-clean-install.md +++ b/docs/tasks/T27-21-deb-build-and-clean-install.md @@ -1,6 +1,6 @@ # T27-21 — Build the DEB from the candidate commit, inspect its metadata, install it clean and run a CLI round trip -> **RP:** RP-01 · **Priority:** P0 · **Status:** IN PROGRESS +> **RP:** RP-01 · **Priority:** P0 · **Status:** SMOKE TESTED > **Board row:** [EXECUTION-BOARD.md §6](../release/EXECUTION-BOARD.md) > **Depends on:** T27-20 @@ -18,7 +18,7 @@ and a CLI import → export round trip produces a correct output file. **Verification:** `dpkg-deb --info dist/native/firecraft_0.2.5_amd64.deb && apt install ./firecraft_0.2.5_amd64.deb` -**Evidence:** Prior clean-container smoke passed 2026-11-06 (matrix); the pipeline's DEB install verification ran stages 1–4b green on `4273471` (run `85502100`) and stage 5 (GUI launch) failed on the undeclared `libxkbcommon-x11-0` — decision 0091 declares the run-time set and the CI installs exactly it +**Evidence:** the DEB builds from the candidate commit inside the release workflow's Debian 12 container (`rust:1.99-slim-bookworm`), `dpkg-deb` reads the package back and the gate fails closed when the desktop entry or the icon is missing from it. The workflow's own package verification then ran against that package in a clean container and **all eight stages passed** (run 44188107435, `build artifacts` job, 2026-11-17T22:24:48Z → 22:25:14Z): install, desktop entry, icon, CLI, CLI document round trip, GUI launch under Xvfb, **upgrade**, uninstall and reinstall. The separate distro matrix installs the same bytes with `apt` on Ubuntu 24.04 and Debian 12 and passes stages 1–5, 7 and 8 ([LINUX-COMPAT.md](../release/LINUX-COMPAT.md)). The upgrade stage is exercised by the workflow's verification, which the matrix does not supply. **Done:** 2026-11-17: the workflow's eight-stage package verification passed on the rehearsal bytes. The pipeline now performs this check on every run (`packaging/deb/verify-install.sh`, stage by stage). On 2026-11-11 the run for @@ -29,7 +29,7 @@ because the package declared only `libc6` and the verification image installed o `libxkbcommon0`. Decision 0091 declares the run-time library set the binary `dlopen`s, and the verification now installs exactly that set plus its harness. -**Blockers:** Clean Debian 12 environment +**Blockers:** — **Stopping condition:** the DEB is built from the candidate commit, installed clean, the round trip passes, and the artifact's digest is recorded. diff --git a/docs/tasks/T27-22-deb-lifecycle-and-desktop-entry.md b/docs/tasks/T27-22-deb-lifecycle-and-desktop-entry.md index ae077975..cc93a68d 100644 --- a/docs/tasks/T27-22-deb-lifecycle-and-desktop-entry.md +++ b/docs/tasks/T27-22-deb-lifecycle-and-desktop-entry.md @@ -21,9 +21,9 @@ installed binary. **Verification:** `dpkg -l; dpkg -P; dpkg-deb --contents … | grep -E 'desktop|icons'` -**Evidence:** The row's premise is fixed in the tree. `build_deb` stages the desktop entry and the icon from the in-tree originals — `usr/share/applications/firecraft.desktop` and `usr/share/icons/hicolor/256x256/apps/firecraft.png` (`xtask/src/dist.rs:973`) — and the gate reads the built package back with `dpkg-deb --contents` and fails closed when either is missing. **The exercise against the fixed package has not been run**: [`packaging/deb/verify-install.sh`](https://github.com/aliasfoxkde/FireCraft/blob/main/packaging/deb/verify-install.sh) carries the eight stages (install, desktop entry, icon, CLI, CLI document round trip, GUI launch, upgrade, uninstall, reinstall), and the recorded negative control is the *published* v0.2.3 package, which installs cleanly and then fails stage 2 — no desktop entry (audit §11.6). The defect is fixed; the reinstall / remove / purge receipt against a package that carries the desktop entry does not exist. **Done:** 2026-11-11: the desktop entry and the icon landed in the DEB payload (PR #107); the clean-container receipt is outstanding. +**Evidence:** The row's premise is fixed in the tree. `build_deb` stages the desktop entry and the icon from the in-tree originals — `usr/share/applications/firecraft.desktop` and `usr/share/icons/hicolor/256x256/apps/firecraft.png` (`xtask/src/dist.rs:973`) — and the gate reads the built package back with `dpkg-deb --contents` and fails closed when either is missing. **The exercise against the fixed package has not been run**: [`packaging/deb/verify-install.sh`](https://github.com/aliasfoxkde/FireCraft/blob/main/packaging/deb/verify-install.sh) carries the eight stages (install, desktop entry, icon, CLI, CLI document round trip, GUI launch, upgrade, uninstall, reinstall), and the recorded negative control is the *published* v0.2.3 package, which installs cleanly and then fails stage 2 — no desktop entry (audit §11.6). The defect is fixed. **The clean-container receipt against a package that carries the desktop entry now exists** (2026-11-17): the eight-stage run passes on Ubuntu 24.04.5 LTS and Debian 12 — stage 2 `PASS: desktop entry installs and names firecraft / firecraft`, `PASS: desktop-file-validate accepts the entry`; stage 3 `PASS: icon installed at /usr/share/icons/hicolor/256x256/apps/firecraft.png`; stage 7 `PASS: removal took the binaries, the desktop entry and the icon with it`; stage 8 `PASS: reinstall restored the payload and the CLI runs`. The row stays `IMPLEMENTED` rather than closing, because **purge is not separately exercised** — stage 7 is an uninstall, and no `dpkg -P` run is in the receipt. **Done:** 2026-11-17: the clean-container eight-stage receipt exists on Ubuntu 24.04 and Debian 12 for a package that carries the desktop entry and the icon. -**Blockers:** the clean-container eight-stage run against a package built from the fixed tree (the CI image and a built .deb) +**Blockers:** purge (`dpkg -P`) is not exercised by the eight-stage script; a purge receipt is still outstanding **Stopping condition:** the package contents include the desktop entry and icons, and the install → reinstall → remove → purge sequence passes with the outputs recorded. diff --git a/docs/tasks/T27-25-pr8-full-suite-and-smoke.md b/docs/tasks/T27-25-pr8-full-suite-and-smoke.md index 501a37d3..25ab2b7d 100644 --- a/docs/tasks/T27-25-pr8-full-suite-and-smoke.md +++ b/docs/tasks/T27-25-pr8-full-suite-and-smoke.md @@ -1,6 +1,6 @@ # T27-25 — Run the full application suite, the CI gate and the web smoke for PR #8 -> **RP:** Phase-2 (PR #8 remediation) · **Priority:** P0 · **Status:** NOT STARTED +> **RP:** Phase-2 (PR #8 remediation) · **Priority:** P0 · **Status:** SMOKE TESTED > **Board row:** [EXECUTION-BOARD.md §6](../release/EXECUTION-BOARD.md) > **Depends on:** T27-23, T27-24 @@ -19,9 +19,9 @@ smoke all pass on the PR head, and the results are recorded with the SHA. **Verification:** `cargo run -p xtask -- ci && cargo run -p xtask -- dist --web` -**Evidence:** none yet — PR #8 was closed unmerged; its review findings are re-hosted in issue #47 and the parity gate in closed issue #86. +**Evidence:** the whole suite ran on the head `f710c524` and passed. PR #8 itself was closed unmerged on 2026-11-12, so the head that carries its re-hosted findings (PR #171, merged as `d3e45cf9`) is the branch head `f710c524`, and that is what was run. **GitHub**, 13 check runs, all terminal on that SHA (the PR run is [44194696167](https://github.com/aliasfoxkde/FireCraft/actions/runs/44194696167)): the two `gate` legs — each `cargo run --package xtask -- ci-no-fmt`, the canonical gate (format, clippy with `-D warnings`, the workspace test suite, layer enforcement, the wasm build) — both `success`; the two `fmt` legs `success`; the two `classify-change` legs `success`; the two `docs-only-check` legs `skipped` because this change is not documentation-only, and the Aegis secret scan runs in its own legs (`cargo run --locked --package xtask -- aegis`), which are `success`; CodeQL `success`; CodeRabbit `success`; the three `Analyze` legs `success`. **GitForge mirror**, pipeline run `65de365f` on the same SHA: all 12 jobs `succeeded` — `clippy-test` (the application test suite), `wasm-web` (`[xtask] wasm ok`), `release` (`[xtask] dist web: 6 file(s), wasm 15783798 B (warning 25000000, hard ceiling 50000000), manifest MANIFEST.sha256, smoke green` and `[xtask] dist --web ok` — the browser smoke against the built bundle), `aegis`, `layers`, `fmt-layers`, `corpus-golden`, `fuzz`, `perf-budgets`, `generated-docs` and `release-ledger` (the committed ledger has no problems). **Done:** 2026-11-17: the full gate, the application test suite, the web build and the browser smoke all green on the head `f710c524`. -**Blockers:** T27-23, T27-24 +**Blockers:** — **Stopping condition:** all four checks pass on the PR head SHA, or every failure is recorded with its cause and owner. diff --git a/docs/tasks/T27-33-second-distro-and-display-check.md b/docs/tasks/T27-33-second-distro-and-display-check.md index 45701e05..6e5751a5 100644 --- a/docs/tasks/T27-33-second-distro-and-display-check.md +++ b/docs/tasks/T27-33-second-distro-and-display-check.md @@ -1,6 +1,6 @@ # T27-33 — Test the DEB and tar.gz on a second distro family and check X11/Wayland, desktop integration and fonts -> **RP:** RP-01 · **Priority:** P1 · **Status:** NOT STARTED +> **RP:** RP-01 · **Priority:** P1 · **Status:** SMOKE TESTED > **Board row:** [EXECUTION-BOARD.md §6](../release/EXECUTION-BOARD.md) > **Depends on:** T27-21 @@ -20,9 +20,9 @@ distribution and version. **Verification:** `docker run bash -lc 'apt install ./firecraft_*.deb && firecraft --version'` -**Evidence:** only Debian 12 has been exercised for the DEB — the rehearsal receipt `docs/release/RECEIPT-0.2.5-DRYRUN-3.md` records the Debian-family DEB build; no Fedora-family image exists on this fleet. +**Evidence:** Six cells, all passing, on 2026-11-17: Ubuntu 24.04.5 LTS, Debian 12 (bookworm) and Fedora 45, each under X11 and under Wayland. The DEB installs with apt on both Debian-family cells and passes the repository's own eight-stage package verification — metadata, desktop entry accepted by `desktop-file-validate`, icon installed, CLI, CLI document round trip (`firecraft-cli export` writes a 97 B PNG with the signature `89504e470d0a1a0a`), GUI launch under Xvfb with exit 124, uninstall and reinstall. The desktop tar.gz launches on all six cells and `fc-list` enumerates fonts on every cell. The harness is in-tree: `packaging/linux/leg-debian-family.sh`, `packaging/linux/leg-fedora-family.sh`, `packaging/linux/leg-wayland.sh` and [`scripts/release-matrix.sh`](https://github.com/aliasfoxkde/FireCraft/blob/main/scripts/release-matrix.sh). The Fedora cells record that a `.deb` is not that family's package format, so the tar.gz is the artifact under test there. The results are published in [LINUX-COMPAT.md](LINUX-COMPAT.md). **Done:** 2026-11-17: six-cell matrix green on Ubuntu 24.04, Debian 12 and Fedora 45. -**Blockers:** no Fedora-family image is available locally +**Blockers:** — **Stopping condition:** a matrix of distro × display results exists, and the gaps are listed explicitly rather than inferred. diff --git a/docs/tasks/T27-34-linux-compatibility-table.md b/docs/tasks/T27-34-linux-compatibility-table.md index 821c5edf..5a3d84e4 100644 --- a/docs/tasks/T27-34-linux-compatibility-table.md +++ b/docs/tasks/T27-34-linux-compatibility-table.md @@ -1,6 +1,6 @@ # T27-34 — Publish a Linux platform compatibility table -> **RP:** RP-01 · **Priority:** P1 · **Status:** NOT STARTED +> **RP:** RP-01 · **Priority:** P1 · **Status:** IMPLEMENTED > **Board row:** [EXECUTION-BOARD.md §6](../release/EXECUTION-BOARD.md) > **Depends on:** T27-33 @@ -19,9 +19,9 @@ untested combination; a reader can tell what is supported from what is merely bu **Verification:** `test -f docs/release/LINUX-COMPAT.md` -**Evidence:** none yet. +**Evidence:** [LINUX-COMPAT.md](LINUX-COMPAT.md) is published and linked from `docs/release/ARTIFACT-MATRIX.md`. It lists the three tested distributions with their versions and architecture, the artifact under test in each cell with its bytes and SHA-256, the formats with no tested row (AppImage and Flatpak — never built; RPM — deferred), and the untested combinations explicitly: every other distribution and release, every architecture but x86_64, real GPUs and vendor drivers, real desktop sessions, session-level desktop integration, the DEB's upgrade path, and the published v0.2.5 assets. **Done:** 2026-11-17: table published and linked. -**Blockers:** T27-33 +**Blockers:** — **Stopping condition:** the table exists, is linked from the artifact matrix, and its "untested" column is non-empty where testing was not performed. diff --git a/docs/tasks/T27-49-bridge-smoke-on-release-bytes.md b/docs/tasks/T27-49-bridge-smoke-on-release-bytes.md index a9c2784d..f60f7c7a 100644 --- a/docs/tasks/T27-49-bridge-smoke-on-release-bytes.md +++ b/docs/tasks/T27-49-bridge-smoke-on-release-bytes.md @@ -1,6 +1,6 @@ # T27-49 — Run the bridge smoke against the release bytes with no skip -> **RP:** RP-05 · **Priority:** P0 · **Status:** BUILD VERIFIED +> **RP:** RP-05 · **Priority:** P0 · **Status:** SMOKE TESTED > **Board row:** [EXECUTION-BOARD.md §6](../release/EXECUTION-BOARD.md) > **Depends on:** T27-48 @@ -20,7 +20,7 @@ verdict is read from the rendered DOM. **Verification:** `cargo run -p xtask -- dist --web` (the bridge-smoke stage). -**Evidence:** The smoke runs and is not skipped. `dist --web` calls `smoke_bridge` ([`xtask/src/dist.rs`](https://github.com/aliasfoxkde/FireCraft/blob/main/xtask/src/dist.rs), line 65), the release workflow installs headless Chrome and fails closed without it (`.github/workflows/release.yml:106`), and `FIRECRAFT_DIST_SKIP_SMOKE` appears nowhere in `.gitforge.yml` or the GitHub workflows — only as a comment at `xtask/src/dist.rs:52`. The rehearsal's `build artifacts` job (run 42937560976) therefore ran the smoke against the tagged bundle and succeeded, and the GitForge release job logs `… smoke green` and `[xtask] dist --web ok` — most recently job `ecb488d1-8965-443f-8309-71405883d754` for commit `fcfe6db1`, finished 2026-11-12T06:18:54Z; twelve release jobs carry that line. **Done:** 2026-11-12: the bridge smoke ran green against the built bundle. +**Evidence:** the smoke runs and is not skipped. `dist --web` calls `smoke_bridge`, the release workflow installs headless Chrome and fails closed without it, and `FIRECRAFT_DIST_SKIP_SMOKE` appears nowhere in `.gitforge.yml` or the GitHub workflows — only as a comment at `xtask/src/dist.rs:52`. The rehearsal's `build artifacts` job (run 44188107435) built the bundle and ran the smoke against it: the job log reads `[xtask] dist web: 7 file(s), wasm 15780814 B …, smoke green` and `[xtask] dist --web ok` at 2026-11-17T22:31:41Z. The GitForge `release` job logs the same line for the mirror builds. Receipt: [RECEIPT-0.2.5-DRYRUN-6.md](../release/RECEIPT-0.2.5-DRYRUN-6.md). **Done:** 2026-11-17: the smoke ran green against the rehearsal's bundle (run 44188107435). **Blockers:** — diff --git a/docs/tasks/T27-50-deployment-reconciliation.md b/docs/tasks/T27-50-deployment-reconciliation.md index a64ba524..e03fc925 100644 --- a/docs/tasks/T27-50-deployment-reconciliation.md +++ b/docs/tasks/T27-50-deployment-reconciliation.md @@ -1,6 +1,6 @@ # T27-50 — Reconcile the production Pages deployment against the release bundle -> **RP:** RP-05 · **Priority:** P0 · **Status:** IMPLEMENTED +> **RP:** RP-05 · **Priority:** P0 · **Status:** SMOKE TESTED > **Board row:** [EXECUTION-BOARD.md §6](../release/EXECUTION-BOARD.md) > **Depends on:** T27-48 @@ -20,9 +20,9 @@ machine-readable report that states "match" or the exact per-file differences. **Verification:** `curl -s https://firecraft-web.pages.dev/MANIFEST.sha256` -**Evidence:** the comparison is a tracked tool and a machine-readable report. [`scripts/reconcile_deployment.py`](https://github.com/aliasfoxkde/FireCraft/blob/main/scripts/reconcile_deployment.py) fetches the deployment's `MANIFEST.sha256` and every file it names, hashes the bytes it actually received, and compares them file by file with the bundle's manifest; [`scripts/reconcile_deployment-selftest.py`](https://github.com/aliasfoxkde/FireCraft/blob/main/scripts/reconcile_deployment-selftest.py) covers the four ways the two can disagree — identical, changed bytes, a deployment-only file, and a bundle file the deployment does not serve. The report against the live deployment is [`docs/release/DEPLOYMENT-RECONCILIATION.json`](https://github.com/aliasfoxkde/FireCraft/blob/main/docs/release/DEPLOYMENT-RECONCILIATION.json) and its result is `differences`: the live deployment serves bundle `2274ac8c2925b61d` while the v0.2.5 release asset carries `f9dbe2f14305b2d9`, each asset name appearing only on its own side of the inventory; `appinfo.json` matches, and `index.html` and `host.html` differ only by the substituted asset names, their integrity digests and the Pages analytics beacon. This confirms the earlier hand comparison (PR #127, `8d5c442`) in machine-readable form: the live deployment corresponds to no identifiable release (blockers B-6, B-11, B-12). **Done:** 2026-11-16: the comparison landed as a tracked tool and a machine-readable report. +**Evidence:** the comparison is a tracked tool and a machine-readable report. `scripts/reconcile_deployment.py` fetches the deployment's `MANIFEST.sha256` and every file it names, hashes the bytes it actually received, and compares them file by file with the bundle's manifest; `scripts/reconcile_deployment-selftest.py` covers the four ways the two can disagree. The report against the live deployment is [`docs/release/DEPLOYMENT-RECONCILIATION.json`](https://github.com/aliasfoxkde/FireCraft/blob/main/docs/release/DEPLOYMENT-RECONCILIATION.json) and its result is `differences`: the live deployment serves bundle `2274ac8c2925b61d` while the v0.2.5 release asset carries `f9dbe2f14305b2d9`, each asset name appearing only on its own side of the inventory; `appinfo.json` matches, and `index.html` and `host.html` differ only by the substituted asset names, their integrity digests and the Pages analytics beacon. **Re-verified 2026-11-17**: the deployment's manifest digest is unchanged at `61e3aeda…` and the comparison against the rehearsal's web bundle reports eight differences — the live bundle has not moved and still corresponds to no identifiable release. **Done:** 2026-11-17: the comparison re-run; the live deployment is unchanged and unreconciled. -**Blockers:** no `CLOUDFLARE_API_TOKEN`; the deploy step in `release.yml` is commented out — a deployment that names a commit stays maintainer-gated +**Blockers:** no `CLOUDFLARE_API_TOKEN`; the deploy step in `release.yml` is commented out — a deployment that names a commit stays maintainer-gated (B-11, B-12) **Stopping condition:** the report shows a byte-level match between the deployed bundle and a tagged release bundle, or it names every differing file and the deployment is diff --git a/docs/tasks/T27-51-cli-archive.md b/docs/tasks/T27-51-cli-archive.md index 95449fc9..036cf714 100644 --- a/docs/tasks/T27-51-cli-archive.md +++ b/docs/tasks/T27-51-cli-archive.md @@ -1,6 +1,6 @@ # T27-51 — Build and verify the standalone CLI archive -> **RP:** RP-05 · **Priority:** P0 · **Status:** IN PROGRESS +> **RP:** RP-05 · **Priority:** P0 · **Status:** SMOKE TESTED > **Board row:** [EXECUTION-BOARD.md §6](../release/EXECUTION-BOARD.md) > **Depends on:** T27-06 @@ -19,7 +19,7 @@ trip produces a correct file; the archive extracts cleanly; the checksum is reco **Verification:** `firecraft-cli --version && firecraft-cli import … && firecraft-cli export …` -**Evidence:** the matrix records the CLI tar.gz as built and manifested; rehearsal `v0.2.5-dryrun.3` (run 42937560976) published `firecraft-cli-0.2.5-dryrun.3-linux-x86_64.tar.gz` (8,109,224 B, sha256 `f646feea…`) as a draft asset and the extracted binary launches. **The archive cannot attest its own build**: the CLI had no `--version`, so the receipt had nothing to say which build it was. `firecraft-cli --version` now prints the stamped version — the archive must be rebuilt for this row to close. **Done:** 2026-11-11: rehearsal receipt written; the rebuilt archive is outstanding. +**Evidence:** the rehearsal published `firecraft-cli-0.2.5-linux-x86_64.tar.gz` (8,158,124 B, sha256 `953633e4…`) as a draft asset (run 44188107435). The archive was downloaded from the draft, re-hashed locally (the digest matches the API's) and extracted, and **the extracted binary attests its own build**: `firecraft-cli --version` prints `firecraft-cli 0.2.5` and exits 0. The distro matrix runs the CLI on all six cells ([LINUX-COMPAT.md](../release/LINUX-COMPAT.md)). The archive is byte-identical across two independent rehearsals. **Done:** 2026-11-17: the published CLI archive attests its own version. **Blockers:** — diff --git a/docs/tasks/T27-52-source-archive.md b/docs/tasks/T27-52-source-archive.md index 385dc049..b092817e 100644 --- a/docs/tasks/T27-52-source-archive.md +++ b/docs/tasks/T27-52-source-archive.md @@ -1,6 +1,6 @@ # T27-52 — Generate the source archive from the exact tagged commit and clean-build it -> **RP:** RP-05 · **Priority:** P0 · **Status:** IN PROGRESS +> **RP:** RP-05 · **Priority:** P0 · **Status:** SMOKE TESTED > **Board row:** [EXECUTION-BOARD.md §6](../release/EXECUTION-BOARD.md) > **Depends on:** T27-06 @@ -20,7 +20,7 @@ tree and any secret; a clean extraction runs `cargo metadata --locked` and **Verification:** `cargo run -p xtask -- dist --source` -**Evidence:** the generator is implemented and wired into `release.yml`; **exercised — and it failed**. The rehearsal's archive (`FireCraft-0.2.5-dryrun.3-source.tar.gz`, asset 733682287) was downloaded from the draft and extracted: `cargo metadata --locked` exits **101** ("cannot update the lock file … because --locked was passed to prevent this"). The manifests are stamped; the lock's dependency *references* and the path dependency's version requirement were not. Fixed by the `xtask stamp` verb and the `xtask stamps` drift gate; the row closes when a rebuilt archive clean-builds. **Done:** 2026-11-11: defect reproduced from the published bytes; fix implemented, verification outstanding. +**Evidence:** the generator is implemented and wired into `release.yml`, and the defect the rehearsal-3 archive exposed — a lockfile whose dependency references did not name the stamped version — was fixed by the `xtask stamp` verb and the `xtask stamps` drift gate. The rehearsal published `firecraft-0.2.5-source.tar.gz` (3,176,024 B, sha256 `0cbee184…`); the workflow's own step reported `clean-extract + cargo metadata --locked verified`, and the archive was then downloaded from the draft and extracted on a **second host**: `cargo metadata --locked` succeeds and `cargo build --locked` finished the dev profile in 53.78s on cargo 1.99.1. The archive's only `.claude/` entry is `.allow_config_edits` — the scratch helper that rehearsal 5 published (`c396b18a`) is gone. The source archive is generated from the tagged commit, carries the sources, the licences and the build inputs, and excludes build outputs and secrets. Receipt: [RECEIPT-0.2.5-DRYRUN-6.md](../release/RECEIPT-0.2.5-DRYRUN-6.md). **Done:** 2026-11-17: the published source archive clean-builds on a second host. **Blockers:** — diff --git a/docs/tasks/T27-53-asset-naming-and-checksums.md b/docs/tasks/T27-53-asset-naming-and-checksums.md index d28df2ee..5909c378 100644 --- a/docs/tasks/T27-53-asset-naming-and-checksums.md +++ b/docs/tasks/T27-53-asset-naming-and-checksums.md @@ -21,7 +21,7 @@ provenance and signature files when they exist; the list verifies with `sha256su **Verification:** `cd dist && sha256sum -c RELEASE_SHA256.txt` -**Evidence:** [`release.yml`](https://github.com/aliasfoxkde/FireCraft/blob/main/.github/workflows/release.yml) generates `RELEASE_SHA256.txt` for 7 files (no provenance/signature files). The rehearsal published it (718 B, sha256 `6ae766e7…`); downloaded from the draft, its 7 entries match the API digests of the 7 assets it covers — 7/7. **Naming is not yet canonical**: the source archive still ships as `FireCraft--source.tar.gz` while every other asset is lowercase; the rename to `firecraft--source.tar.gz` is implemented and lands with this row's next rebuild **Done:** 2026-11-11: checksum list verified against the published digests (rehearsal `v0.2.5-dryrun.3`). +**Evidence:** [`release.yml`](https://github.com/aliasfoxkde/FireCraft/blob/main/.github/workflows/release.yml) generates `RELEASE_SHA256.txt` for 7 files and names each file explicitly, because the globs over-match. The naming defect the rehearsal-3 receipt found is closed: every published asset now carries the lowercase canonical name (`firecraft-0.2.5-source.tar.gz`, not `FireCraft-…`). The rehearsal published the list (664 B, sha256 `af054b99…`); downloaded from the draft, its 7 entries match the API digests of the 7 assets it covers — 7/7. The list does not cover itself and does not cover the manifest, which is why it holds seven entries for nine assets. The native `MANIFEST.sha256` is byte-identical across two rehearsals. **Done:** 2026-11-17: the checksum list verified 7/7 against the rehearsal's published digests. **Blockers:** — diff --git a/docs/tasks/T27-54-release-manifest.md b/docs/tasks/T27-54-release-manifest.md index 6c67b76a..73e1c8f6 100644 --- a/docs/tasks/T27-54-release-manifest.md +++ b/docs/tasks/T27-54-release-manifest.md @@ -1,6 +1,6 @@ # T27-54 — Generate and verify a machine-readable release manifest -> **RP:** RP-06 · **Priority:** P0 · **Status:** BUILD VERIFIED(validation pending) +> **RP:** RP-06 · **Priority:** P0 · **Status:** SMOKE TESTED(validation pending) > **Board row:** [EXECUTION-BOARD.md §6](../release/EXECUTION-BOARD.md) > **Depends on:** T27-53 > **Decision:** [0085](../decisions/0089-release-manifest-shape.md) · **Schema:** [RELEASE-MANIFEST.md](../release/RELEASE-MANIFEST.md) @@ -21,9 +21,9 @@ against the actual files. **Verification:** `jq -r '.version,.artifacts[0].sha256' dist/release-manifest.json` -**Evidence:** implemented on `feat/release-manifest` (PR #106, merged in the train as `658f015`) per [decision 0089](https://github.com/aliasfoxkde/FireCraft/blob/main/docs/decisions/0089-release-manifest-shape.md): `xtask dist --manifest` writes `dist/release-manifest.json` (schema `firecraft.release-manifest/1`) for the eight published assets — canonical name, path, size, SHA-256, format, target triple and verification status — plus the run's version, tag, source revision, run URL and timestamp; `dist --verify-manifest` re-checks it against the files; `release.yml` publishes it as a ninth asset and validates it against the downloads. Unit tests cover the shape, the determinism and every rejection path. **No CI receipt and no published manifest yet** — the row closes when a rehearsal publishes one. **Done:** 2026-11-11: implementation merged (PR #106 → #107). +**Evidence:** `xtask dist --manifest` writes `dist/release-manifest.json` (schema `firecraft.release-manifest/1`) with one entry per published artifact — canonical name, path, size, SHA-256, format, target triple and verification status — plus the run's version, **tag**, source revision, **run URL** and timestamp; `dist --verify-manifest` re-checks it against the files. The rehearsal published the manifest (3,295 B, sha256 `54b8a231…`) and the workflow's aggregate gate validated it against the downloaded assets. **The manifest names its tag and its run**: `tag` `v0.2.5-dryrun.6`, `source_sha` `c396b18afba0a…`, `run_url` `https://github.com/aliasfoxkde/FireCraft/actions/runs/44188107435`, `run_id` `44188107435`, `built_at` `2026-11-17T22:31:44Z`, 8 artifact entries. Both gaps that made that impossible were closed first — the caller now states the tag and `xtask` fails closed when the stated tag does not name the workspace version, and the run URL is reconstructed from `GITHUB_SERVER_URL`, `GITHUB_REPOSITORY` and `GITHUB_RUN_ID` because GitHub Actions exports no `GITHUB_RUN_URL`. **Done:** 2026-11-17: a published manifest that names its tag and its run. -**Blockers:** T27-53's canonical naming is still `NOT STARTED`; the manifest adopts the names the release already enforces, and its asset table moves with T27-53 (decision 0089) +**Blockers:** — **Stopping condition:** the manifest exists, verifies, and is published alongside the assets. diff --git a/docs/tasks/T27-55-sbom-regenerate-and-validate.md b/docs/tasks/T27-55-sbom-regenerate-and-validate.md index ce61c9e9..1c359857 100644 --- a/docs/tasks/T27-55-sbom-regenerate-and-validate.md +++ b/docs/tasks/T27-55-sbom-regenerate-and-validate.md @@ -1,6 +1,6 @@ # T27-55 — Regenerate and validate the CycloneDX SBOM -> **RP:** RP-06 · **Priority:** P0 · **Status:** BUILD VERIFIED +> **RP:** RP-06 · **Priority:** P0 · **Status:** SMOKE TESTED > **Board row:** [EXECUTION-BOARD.md §6](../release/EXECUTION-BOARD.md) > **Depends on:** T27-06 @@ -19,7 +19,7 @@ against the CycloneDX 1.5 schema; it is compared component-by-component against **Verification:** `cargo run -p xtask -- dist --sbom && jq .metadata.component dist/sbom/firecraft-0.2.5-sbom.cdx.json` -**Evidence:** the CycloneDX 1.5 schema and the SPDX sibling it points at are vendored under `xtask/schema/`, and `xtask dist --sbom` validates the document it builds against them *before* writing it and compares its component inventory with `Cargo.lock`'s `[[package]]` sections, naming both sides of any difference (decision [0119](../decisions/0119-the-sbom-is-validated-against-the-schema-it-claims.md)). The validator refuses a schema that uses a keyword it does not implement, so a partial implementation cannot report a bad document as valid. The generator's licence field was wrong for every crate whose licence is an SPDX expression — `MIT OR Apache-2.0` is an expression, not an identifier — and the identifier list is now read from the schema itself. The comparison then failed on its first real run — the CI release job reported `in Cargo.lock but not in the SBOM: glob 0.3.4 / libwebp-sys 0.14.4` — because the inventory came from `cargo metadata`'s *resolved* graph (560 packages) while the lockfile *pins* 562: Cargo resolves the lockfile independently of any one feature selection, so optional and platform-specific crates that nothing here enables are pinned but not resolved. The generator now takes the lockfile as the inventory, lists the two pinned-but-unlinked packages as components with an empty dependency edge (pinned, not linked) and the archive digest the lockfile records, and the comparison is a real assertion rather than a tautology against a resolved graph. **Outstanding**: publishing the newly validated SBOM and verifying the download belongs to the next release rehearsal **Done:** 2026-11-16: schema validation and the `Cargo.lock` comparison implemented and verified, including the lockfile-inventory fix the first CI run exposed. +**Evidence:** the CycloneDX 1.5 schema and the SPDX sibling it points at are vendored under `xtask/schema/`, and `xtask dist --sbom` validates the document it builds against them *before* writing it and compares its component inventory with `Cargo.lock`'s `[[package]]` sections, naming both sides of any difference (decision [0119](../decisions/0119-the-sbom-is-validated-against-the-schema-it-claims.md)). The inventory is the lockfile's own, not a resolved graph, so the two pinned-but-unlinked packages are listed as components with an empty dependency edge. The rehearsal published the SBOM (369,745 B, sha256 `e50516b2…`): the workflow's step reported `562 components, 562 locked package(s), schema-valid`, and the aggregate gate validated the download. Two independent rehearsals' SBOMs are identical in every field except `metadata.timestamp` — same components, same dependency graph, same serial number. **Done:** 2026-11-17: the validated SBOM published and verified. **Blockers:** — diff --git a/docs/tasks/T27-57-workflow-hardening.md b/docs/tasks/T27-57-workflow-hardening.md index c6fcf3e6..805e90d0 100644 --- a/docs/tasks/T27-57-workflow-hardening.md +++ b/docs/tasks/T27-57-workflow-hardening.md @@ -1,6 +1,6 @@ # T27-57 — Harden the release workflows -> **RP:** RP-06 · **Priority:** P1 · **Status:** BUILD VERIFIED +> **RP:** RP-06 · **Priority:** P1 · **Status:** SMOKE TESTED > **Board row:** [EXECUTION-BOARD.md §6](../release/EXECUTION-BOARD.md) > **Depends on:** T27-18 @@ -107,7 +107,7 @@ flag, so it does not yet demonstrate this mode. The three self-tests and the pin check are in the GitHub backup CI; mirroring them into the GitForge pipeline waits for the `:2` pipeline shape (PR #123, PR #129). -**Evidence:** **Pinning implemented 2026-11-11** (decision [0099](../decisions/0099-workflow-actions-are-pinned.md)): all 19 action references in `ci.yml`, `release.yml` and `deploy-site.yml` name a 40-hex commit SHA with the version in a trailing comment; `scripts/validate_workflow_pins.py` fails the CI on an unpinned reference, and `scripts/validate_workflow_pins-selftest.sh` demonstrates each failure mode with one deliberately broken input. Least-privilege permissions were already declared (`contents: read` at workflow level, `contents: write` only for the draft-release job) and the aggregate gate already fails closed on a missing, empty, misnamed or digest-mismatched artifact. The deploy trigger no longer includes the workflow's own file, so a workflow edit is not a production publication. **Both remaining criteria implemented 2026-11-12** (decision [0100](../decisions/0100-release-mode-classification-and-a-tested-release-gate.md)): the aggregate gate moved out of inline YAML into `scripts/validate_release_aggregate.sh`, which now also rejects an artifact the release does not publish and requires every required platform (`x86_64-unknown-linux-gnu`, `wasm32-unknown-unknown`) to be represented in the release manifest; `scripts/validate_release_aggregate-selftest.sh` breaks a complete artifact set one thing at a time — missing, empty, extra artifact, checksum mismatch, manifest version mismatch, manifest digest mismatch, an unlisted asset, an unknown schema, a skipped platform, and a platform the manifest does not represent. A safe test-release mode exists: `scripts/classify_release_mode.sh` maps `vX.Y.Z` to production and `vX.Y.Z-dryrun.N` to a rehearsal and refuses every other shape; a manual `workflow_dispatch` run may only rehearse; a rehearsal release is a prerelease titled `TEST ` whose notes carry a do-not-publish banner, and its disposable tag is created at the commit that built the artifacts. `scripts/classify_release_mode-selftest.sh` demonstrates each refused shape, and `tests/release-assets-selftest.sh` drives the `gh` argument vector against a stub. **Toolchain pinned 2026-11-12** (decision [0102](../decisions/0102-coverage-floor-is-measured-with-a-pinned-toolchain.md)): the workflows install `toolchain: 1.99.1` and the CI image is `rust:1.99.1-slim-bookworm`, because the unpinned `stable` moved the coverage floor silently — rustc 1.100.0 measures a byte-identical Rust tree at 77.0%/76.1% where 1.99.1 measures 81.9%/78.9%, while the GitForge image stayed on 1.99. The gate now names the compiler that measured it. **Outstanding**: the rehearsal path has not been run end to end through `workflow_dispatch` (it needs this workflow on `main`); the manual rehearsal in issue #99 predates the `TEST` title and the prerelease flag **Done:** 2026-11-12: `gate` green on head `002eb56` (PR #130, merged 2026-11-12). +**Evidence:** the pinning, the least-privilege permissions, the fail-closed aggregate gate, the release-mode classifier and the pinned toolchain are all in place and tested (decisions [0099](../decisions/0099-workflow-actions-are-pinned.md), [0100](../decisions/0100-release-mode-classification-and-a-tested-release-gate.md), [0102](../decisions/0102-coverage-floor-is-measured-with-a-pinned-toolchain.md)). **The remaining criterion is met**: the rehearsal path has now been run end to end through `workflow_dispatch` — twice, on runs 44185273916 (`v0.2.5-dryrun.5`) and 44188107435 (`v0.2.5-dryrun.6`), both `completed`/`success`, each producing a `TEST`-titled prerelease draft with the do-not-publish banner, each classified as a rehearsal by `scripts/classify_release_mode.sh`, each passing the aggregate gate. A manual run may only rehearse, and the rehearsal's artifacts are bound to one tag SHA. **Done:** 2026-11-17: the rehearsal path ran end to end through `workflow_dispatch`. **Blockers:** — diff --git a/docs/tasks/T27-60-release-certification-report.md b/docs/tasks/T27-60-release-certification-report.md index fe1c0829..54a0a042 100644 --- a/docs/tasks/T27-60-release-certification-report.md +++ b/docs/tasks/T27-60-release-certification-report.md @@ -1,6 +1,6 @@ # T27-60 — Produce the release certification report for the approved scope -> **RP:** RP-08 · **Priority:** P0 · **Status:** NOT STARTED +> **RP:** RP-08 · **Priority:** P0 · **Status:** SMOKE TESTED > **Board row:** [EXECUTION-BOARD.md §6](../release/EXECUTION-BOARD.md) > **Depends on:** T27-19, T27-50 @@ -24,9 +24,9 @@ the residual risk and the independent verification results. **Verification:** `test -f docs/release/CERTIFICATION.md` -**Evidence:** no certification report exists — the rehearsal receipt `docs/release/RECEIPT-0.2.5-DRYRUN-3.md` certifies nothing: no artifact is signed and no provenance or attestation exists, so the ladder's ceiling for this repository is rung 8. +**Evidence:** the report exists: [[CERTIFICATION.md](../release/CERTIFICATION.md)](../release/CERTIFICATION.md). It names the tag, the source SHA, the workspace version, the approved matrix, every artifact with its size, digest, runner and toolchain, the install/launch/smoke evidence, the signature/SBOM/provenance status, the deployment reconciliation, the CI gates, the limitations, the unsupported targets, the residual risk and the independent verification results. The independent checks are real: `RELEASE_SHA256.txt` downloaded from the draft matches 7 of 7 API digests; three downloaded assets re-hash 3 of 3; the source archive clean-builds on a second host; the native artifacts are byte-identical across two independent rehearsals. **The report certifies nothing beyond that**: nothing is signed and no provenance exists, so no artifact reaches rung 9 of the evidence ladder. **Done:** 2026-11-17: the certification report published in the repository. -**Blockers:** — +**Blockers:** B-14 — no signing identity and no provenance generator exist, so no artifact can be `CERTIFIED`; and approved-scope item 8 (the verified production deployment) is blocked by B-11/B-12 **Stopping condition:** the report exists and every claim in it is backed by a cited receipt. Unsupported targets are listed as unsupported, not omitted. diff --git a/packaging/linux/leg-debian-family.sh b/packaging/linux/leg-debian-family.sh new file mode 100644 index 00000000..4b39fc5a --- /dev/null +++ b/packaging/linux/leg-debian-family.sh @@ -0,0 +1,66 @@ +#!/usr/bin/env bash +# The Debian-family leg of the Linux compatibility matrix (T27-33). +# +# It installs the DEB with apt — so apt itself resolves the package's +# declared dependencies — and then runs the eight-stage package +# verification, and separately unpacks the desktop tar.gz and launches +# it under X11. +# +# Run inside a container with: +# /artifacts the release artifacts (read-only) +# /repo the repository checkout (read-only) +# +# Usage (inside the container): +# bash /repo/packaging/linux/leg-debian-family.sh +set -euo pipefail + +echo "== image: $(grep PRETTY_NAME /etc/os-release)" + +apt-get update -qq >/dev/null +apt-get install -y -qq --no-install-recommends \ + xvfb xauth libgl1 libx11-6 libxcursor1 libxrandr2 libxi6 \ + libxkbcommon0 libxkbcommon-x11-0 desktop-file-utils \ + fonts-dejavu-core ca-certificates fontconfig >/dev/null + +echo "== install the DEB (apt resolves the declared dependencies)" +apt-get install -y -qq /artifacts/firecraft_*.deb >/dev/null + +echo "== the eight-stage package verification" +bash /repo/packaging/deb/verify-install.sh \ + /artifacts/firecraft_*.deb "" /repo/corpus/fixtures/composed.fw.png + +echo "== the desktop tar.gz, unpacked and run" +mkdir -p /opt/firecraft-tar +tar xzf /artifacts/firecraft-[0-9]*-linux-x86_64.tar.gz -C /opt/firecraft-tar +ls -R /opt/firecraft-tar +# The archive's layout is `/bin/firecraft`; a glob states it +# without walking the tree. +TAR_BIN="" +for candidate in /opt/firecraft-tar/*/bin/firecraft; do + if [ -x "$candidate" ]; then + TAR_BIN="$candidate" + break + fi +done +[ -n "$TAR_BIN" ] || { echo "FAIL: the tar.gz carries no firecraft binary" >&2; exit 1; } +set +e +timeout 25 xvfb-run -a "$TAR_BIN" >/tmp/tar-gui.log 2>&1 +TAR_STATUS=$? +set -e +if [ "$TAR_STATUS" = "124" ]; then + echo "PASS: the tar.gz desktop binary launched and stayed up" +else + cat /tmp/tar-gui.log >&2 || true + echo "FAIL: the tar.gz desktop binary exited with status $TAR_STATUS" >&2 + exit 1 +fi + +echo "== fonts: the desktop entry's font dependency" +if command -v fc-list >/dev/null 2>&1; then + fc-list >/dev/null + echo "PASS: fontconfig enumerates fonts" +else + echo "NOTE: fc-list is unavailable" +fi + +echo "== all stages passed" diff --git a/packaging/linux/leg-fedora-family.sh b/packaging/linux/leg-fedora-family.sh new file mode 100644 index 00000000..9456637b --- /dev/null +++ b/packaging/linux/leg-fedora-family.sh @@ -0,0 +1,75 @@ +#!/usr/bin/env bash +# The Fedora-family leg of the Linux compatibility matrix (T27-33). +# +# The DEB is a Debian package format: no package manager on this family +# can install it. That is a fact about the format, not a gap in the +# test, so the leg says so out loud and then covers the *tar.gz* — the +# format-agnostic artifact — on this family instead. +# +# Run inside a container with: +# /artifacts the release artifacts (read-only) +# /repo the repository checkout (read-only) +# +# Usage (inside the container): +# bash /repo/packaging/linux/leg-fedora-family.sh +set -euo pipefail + +echo "== image: $(grep PRETTY_NAME /etc/os-release)" + +dnf install -y -q \ + xorg-x11-server-Xvfb mesa-libGL libXcursor libXrandr libXi \ + libxkbcommon libxkbcommon-x11 mesa-dri-drivers dejavu-sans-fonts \ + fontconfig tar gzip >/dev/null + +echo "== the DEB is not installable on this family" +if command -v dpkg >/dev/null 2>&1; then + echo "NOTE: dpkg is present but a .deb is not this family's package format" +else + echo "NOTE: no dpkg in this image; a .deb cannot be installed on this family" +fi + +echo "== the desktop tar.gz, unpacked and run" +mkdir -p /opt/firecraft-tar +tar xzf /artifacts/firecraft-[0-9]*-linux-x86_64.tar.gz -C /opt/firecraft-tar +ls -R /opt/firecraft-tar +# The archive's layout is `/bin/firecraft`; a glob states it +# without walking the tree. +TAR_BIN="" +for candidate in /opt/firecraft-tar/*/bin/firecraft; do + if [ -x "$candidate" ]; then + TAR_BIN="$candidate" + break + fi +done +[ -n "$TAR_BIN" ] || { echo "FAIL: the tar.gz carries no firecraft binary" >&2; exit 1; } +TAR_CLI="" +for candidate in /opt/firecraft-tar/*/bin/firecraft-cli; do + if [ -x "$candidate" ]; then + TAR_CLI="$candidate" + break + fi +done +if [ -n "$TAR_CLI" ]; then + "$TAR_CLI" --help | head -3 + echo "PASS: the tar.gz CLI runs" +fi +set +e +timeout 25 xvfb-run -a "$TAR_BIN" >/tmp/tar-gui.log 2>&1 +TAR_STATUS=$? +set -e +if [ "$TAR_STATUS" = "124" ]; then + echo "PASS: the tar.gz desktop binary launched and stayed up" +else + cat /tmp/tar-gui.log >&2 || true + echo "FAIL: the tar.gz desktop binary exited with status $TAR_STATUS" >&2 + exit 1 +fi + +if command -v fc-list >/dev/null 2>&1; then + fc-list >/dev/null + echo "PASS: fontconfig enumerates fonts" +else + echo "NOTE: fc-list is unavailable" +fi + +echo "== all stages passed" diff --git a/packaging/linux/leg-wayland.sh b/packaging/linux/leg-wayland.sh new file mode 100644 index 00000000..7d6532de --- /dev/null +++ b/packaging/linux/leg-wayland.sh @@ -0,0 +1,106 @@ +#!/usr/bin/env bash +# The Wayland leg of the Linux compatibility matrix (T27-33). +# +# It runs the desktop binary against a real Wayland compositor (weston's +# headless backend) and proves the window comes up there, not only under +# X11. +# +# The cell is self-contained: it puts a FireCraft binary on the box +# before it launches anything. On a Debian-family image that means +# installing the DEB (the artifact under test); on a Fedora-family image +# it means unpacking the desktop tar.gz, because a .deb is not that +# family's package format. +# +# Run inside a container with: +# /artifacts the release artifacts (read-only) +# /repo the repository checkout (read-only) +# +# Usage (inside the container): +# bash /repo/packaging/linux/leg-wayland.sh +set -euo pipefail + +echo "== image: $(grep PRETTY_NAME /etc/os-release)" + +if command -v apt-get >/dev/null 2>&1; then + # The package lists must be refreshed first: a fresh image carries + # no lists, and `apt-get install` without them answers "Unable to + # locate package" for packages that exist. + apt-get update -qq >/dev/null + apt-get install -y -qq --no-install-recommends \ + weston wayland-protocols libwayland-client0 libwayland-server0 \ + libxkbcommon0 libxkbcommon-x11-0 libgl1 libx11-6 libxcursor1 \ + libxrandr2 libxi6 fonts-dejavu-core fontconfig >/dev/null +else + dnf install -y -q weston libwayland-client libwayland-server \ + libwayland-cursor libxkbcommon libxkbcommon-x11 mesa-libGL \ + libXcursor libXrandr libXi dejavu-sans-fonts fontconfig >/dev/null +fi + +# --- put a desktop binary on the box --------------------------------- +BIN="${FIRECRAFT_BIN:-}" +if [ -z "$BIN" ] || [ ! -x "$BIN" ]; then + if command -v apt-get >/dev/null 2>&1 && ls /artifacts/firecraft_*.deb >/dev/null 2>&1; then + echo "== install the DEB (apt resolves the declared dependencies)" + apt-get install -y -qq /artifacts/firecraft_*.deb >/dev/null + BIN=/usr/bin/firecraft + else + echo "== the tar.gz is this family's artifact; unpack it" + mkdir -p /opt/firecraft-tar + tar xzf /artifacts/firecraft-[0-9]*-linux-x86_64.tar.gz -C /opt/firecraft-tar + # The archive's layout is `/bin/firecraft`; a glob + # states it without walking the tree. + for candidate in /opt/firecraft-tar/*/bin/firecraft; do + if [ -x "$candidate" ]; then + BIN="$candidate" + break + fi + done + fi +fi +[ -n "$BIN" ] && [ -x "$BIN" ] || + { echo "FAIL: no firecraft desktop binary is available to launch" >&2; exit 1; } +echo "== the binary under test: $BIN" + +export XDG_RUNTIME_DIR=/run/firecraft-wayland +mkdir -p "$XDG_RUNTIME_DIR" +chmod 700 "$XDG_RUNTIME_DIR" + +echo "== start the headless Wayland compositor" +weston --backend=headless-backend.so --socket=wayland-0 --width=1280 --height=800 \ + >/tmp/weston.log 2>&1 & +WESTON_PID=$! +for _ in $(seq 1 30); do + [ -S "$XDG_RUNTIME_DIR/wayland-0" ] && break + sleep 1 +done +if [ ! -S "$XDG_RUNTIME_DIR/wayland-0" ]; then + cat /tmp/weston.log >&2 || true + echo "FAIL: the compositor never created its socket" >&2 + kill "$WESTON_PID" 2>/dev/null || true + exit 1 +fi +echo "PASS: weston is serving $XDG_RUNTIME_DIR/wayland-0" + +echo "== launch the desktop binary on Wayland" +set +e +WAYLAND_DISPLAY=wayland-0 timeout 25 "$BIN" >/tmp/wayland-gui.log 2>&1 +WAYLAND_STATUS=$? +set -e +kill "$WESTON_PID" 2>/dev/null || true + +if [ "$WAYLAND_STATUS" = "124" ]; then + echo "PASS: the desktop binary launched on Wayland and stayed up" +else + cat /tmp/wayland-gui.log >&2 || true + echo "FAIL: the desktop binary exited with status $WAYLAND_STATUS on Wayland" >&2 + exit 1 +fi + +if command -v fc-list >/dev/null 2>&1; then + fc-list >/dev/null + echo "PASS: fontconfig enumerates fonts" +else + echo "NOTE: fc-list is unavailable" +fi + +echo "== all stages passed" diff --git a/scripts/release-matrix.sh b/scripts/release-matrix.sh new file mode 100644 index 00000000..0cb3ecbb --- /dev/null +++ b/scripts/release-matrix.sh @@ -0,0 +1,65 @@ +#!/usr/bin/env bash +# The Linux compatibility matrix (T27-33) — three distro families × two +# display servers, six cells, each in its own container. +# +# Usage: scripts/release-matrix.sh [results-dir] +# +# must hold the DEB and the desktop tar.gz, e.g. the +# output of `cargo run -p xtask -- dist --native`. +# +# The leg scripts and the package verifier come from the checkout, so +# the matrix tests the repository's own verification, not a copy of it. +# +# The `:z` on every bind mount is not decoration: on a SELinux-enforcing +# host an unlabelled mount from /var/tmp is unreadable inside the +# container, and the cell fails with "Permission denied" before it can +# run anything. +# +# Exit status is 0 only when every cell passed. +set -uo pipefail + +REPO_ROOT="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)" +ART="${1:?usage: release-matrix.sh [results-dir]}" +RESULTS="${2:-/var/tmp/firecraft-matrix/results}" +mkdir -p "$RESULTS" +: >"$RESULTS/matrix.tsv" + +run_cell() { + local label="$1" image="$2" script="$3" bin="$4" + local log="$RESULTS/$label.log" + echo "== cell: $label ($image)" + local status=0 + docker run --rm \ + -v "$ART:/artifacts:ro,z" \ + -v "$REPO_ROOT:/repo:ro,z" \ + -e FIRECRAFT_BIN="$bin" \ + "$image" bash "/repo/packaging/linux/$script" >"$log" 2>&1 || status=$? + if [ "$status" = 0 ]; then + echo " PASS $label (log: $log)" + else + echo " FAIL $label exit=$status (log: $log)" + tail -40 "$log" | sed 's/^/ | /' + fi + printf '%s\t%s\t%s\n' "$label" "$image" "$status" >>"$RESULTS/matrix.tsv" + return 0 +} + +# --- Debian family: the DEB installs with apt, plus the tar.gz -------- +run_cell ubuntu-24.04-x11 ubuntu:24.04 leg-debian-family.sh /usr/bin/firecraft +run_cell debian-12-x11 debian:12-slim leg-debian-family.sh /usr/bin/firecraft +run_cell ubuntu-24.04-wayland ubuntu:24.04 leg-wayland.sh /usr/bin/firecraft +run_cell debian-12-wayland debian:12-slim leg-wayland.sh /usr/bin/firecraft + +# --- Fedora family: the DEB is not this family's package format, so the +# tar.gz is the artifact under test there. +run_cell fedora-x11 fedora:latest leg-fedora-family.sh "" +run_cell fedora-wayland fedora:latest leg-wayland.sh "" + +echo "== matrix" +column -t -s $'\t' "$RESULTS/matrix.tsv" + +if grep -qv $'\t0$' "$RESULTS/matrix.tsv"; then + echo "== at least one cell failed" >&2 + exit 1 +fi +echo "== every cell passed" diff --git a/xtask/src/dist.rs b/xtask/src/dist.rs index 0a429c1f..40ade18a 100644 --- a/xtask/src/dist.rs +++ b/xtask/src/dist.rs @@ -1465,6 +1465,9 @@ fn parse_recipes() -> Result { "build-flatpak.sh", "build-macos-dmg.sh", "deb/verify-install.sh", + "linux/leg-debian-family.sh", + "linux/leg-fedora-family.sh", + "linux/leg-wayland.sh", ] { let script = workspace_root().join("packaging").join(name); let status = std::process::Command::new("bash") @@ -1480,6 +1483,22 @@ fn parse_recipes() -> Result { } checked += 1; } + // The compatibility-matrix orchestrator lives beside the other + // repository scripts rather than under packaging/, because it drives + // the packaging recipes rather than producing a package. + let matrix = workspace_root().join("scripts/release-matrix.sh"); + let status = std::process::Command::new("bash") + .args(["-n"]) + .arg(&matrix) + .output() + .map_err(|e| format!("spawn bash -n: {e}"))?; + if !status.status.success() { + return Err(format!( + "scripts/release-matrix.sh does not parse:\n{}", + String::from_utf8_lossy(&status.stderr).trim() + )); + } + checked += 1; // The Flatpak manifest is data, not a script: a YAML parse proves // structure. The saphyr dependency lives in xtask's Cargo.toml; // `serde_yaml` was rejected here because it is unmaintained and @@ -2059,20 +2078,100 @@ fn manifest_context(version: &str) -> Result { )); } } - let tag = git_tag.or_else(|| { - env_value("GITHUB_REF_NAME") - .filter(|name| name.strip_prefix('v').unwrap_or(name.as_str()) == version) - }); + let tag = git_tag + .or_else(|| { + env_value("GITHUB_REF_NAME") + .filter(|name| name.strip_prefix('v').unwrap_or(name.as_str()) == version) + }) + .or(stated_release_tag(version)?); Ok(ManifestContext { version: version.to_owned(), tag, source_sha, - run_url: env_value("GITHUB_RUN_URL"), + run_url: run_url(), run_id: env_value("GITHUB_RUN_ID"), built_at: build_timestamp(), }) } +/// The URL of the CI run that produced the artifacts. +/// +/// GitHub Actions does not export `GITHUB_RUN_URL` — the default +/// environment carries the run *id* but not the run *url* — so a +/// manifest generated in CI came out with `"run_url": null` while +/// naming its run id. The URL is reconstructed from the three variables +/// that are exported: the server, the repository and the run id. +/// +/// The reconstruction is pure so it can be tested without the +/// environment. +fn run_url() -> Option { + if let Some(url) = env_value("GITHUB_RUN_URL") { + return Some(url); + } + built_run_url( + env_value("GITHUB_SERVER_URL").as_deref(), + env_value("GITHUB_REPOSITORY").as_deref(), + env_value("GITHUB_RUN_ID").as_deref(), + ) +} + +/// Builds the canonical run URL from the three CI facts, or `None` when +/// any of them is absent. +/// +/// `host` is the API root (`https://github.com`), `repository` is +/// `owner/name`, and `run_id` is the numeric run. +fn built_run_url( + host: Option<&str>, + repository: Option<&str>, + run_id: Option<&str>, +) -> Option { + let host = host?.trim_end_matches('/'); + let repository = repository?; + let run_id = run_id?; + if host.is_empty() || repository.is_empty() || run_id.is_empty() { + return None; + } + Some(format!("{host}/{repository}/actions/runs/{run_id}")) +} + +/// The release tag the caller states for this build, when it states +/// one. +/// +/// A rehearsal is built from `main` and its tag is created *after* the +/// build, so neither the git tag nor the CI ref name can name it: the +/// manifest came out with `"tag": null` while the release it describes +/// had a tag. The caller therefore states the tag, and this reads that +/// statement. +/// +/// The statement is checked, not trusted: it has to name the workspace +/// version, either exactly (`v0.2.5`) or with a disposable suffix +/// (`v0.2.5-dryrun.4`). A tag that names a different version is an +/// error, not a silent omission — the artifacts carry one version and +/// the manifest may not claim another. +fn stated_release_tag(version: &str) -> Result, String> { + let Some(tag) = env_value("RELEASE_TAG") else { + return Ok(None); + }; + if stated_tag_names_version(&tag, version) { + Ok(Some(tag)) + } else { + Err(format!( + "the stated release tag {tag} does not name the workspace version {version} — \ + the tree and the tag disagree about what is being released" + )) + } +} + +/// Whether a stated tag names this workspace version. +/// +/// The rule is pure so it can be exercised without touching the +/// environment: the production tag names the version exactly, and a +/// rehearsal tag names the version and a disposable suffix. +fn stated_tag_names_version(tag: &str, version: &str) -> bool { + let stripped = tag.strip_prefix('v').unwrap_or(tag); + stripped == version || stripped.starts_with(&format!("{version}-")) +} + /// Runs a git command in `root` and returns its trimmed stdout. /// /// `None` means the command produced no output — a commit that carries @@ -2743,8 +2842,8 @@ mod manifest_tests { #![allow(clippy::unwrap_used, clippy::expect_used, clippy::panic)] use super::{ ArtifactSpec, MANIFEST_FILE_NAME, MANIFEST_SCHEMA, ManifestContext, VERIFICATION_COMPUTED, - VERIFICATION_VERIFIED, build_manifest, digest_and_size, parse_checksum_list, release_specs, - render_manifest, verify_manifest_at, + VERIFICATION_VERIFIED, build_manifest, built_run_url, digest_and_size, parse_checksum_list, + release_specs, render_manifest, stated_tag_names_version, verify_manifest_at, }; use std::path::{Path, PathBuf}; @@ -2905,6 +3004,62 @@ mod manifest_tests { ); } + #[test] + fn a_stated_tag_must_name_the_workspace_version() { + // The production tag names the version exactly. + assert!(stated_tag_names_version("v0.2.5", "0.2.5")); + assert!(stated_tag_names_version("0.2.5", "0.2.5")); + // A rehearsal tag names the version and a disposable suffix. + assert!(stated_tag_names_version("v0.2.5-dryrun.4", "0.2.5")); + assert!(stated_tag_names_version("0.2.5-rc.1", "0.2.5")); + // A tag that names another version is refused, and so is one + // that merely starts with the same digits. + assert!(!stated_tag_names_version("v0.2.6", "0.2.5")); + assert!(!stated_tag_names_version("v0.2.50", "0.2.5")); + assert!(!stated_tag_names_version("main", "0.2.5")); + assert!(!stated_tag_names_version("", "0.2.5")); + } + + #[test] + fn the_run_url_is_reconstructed_from_the_ci_facts() { + // GitHub Actions does not export GITHUB_RUN_URL, so the manifest + // has to build the URL from the three variables it does export. + assert_eq!( + built_run_url( + Some("https://github.com"), + Some("aliasfoxkde/FireCraft"), + Some("44179611600") + ), + Some("https://github.com/aliasfoxkde/FireCraft/actions/runs/44179611600".to_owned()) + ); + // A trailing slash on the server root must not double up. + assert_eq!( + built_run_url( + Some("https://github.com/"), + Some("aliasfoxkde/FireCraft"), + Some("44179611600") + ), + Some("https://github.com/aliasfoxkde/FireCraft/actions/runs/44179611600".to_owned()) + ); + // A missing fact leaves the field absent rather than half-built. + assert_eq!( + built_run_url( + Some("https://github.com"), + Some("aliasfoxkde/FireCraft"), + None + ), + None + ); + assert_eq!( + built_run_url(None, Some("aliasfoxkde/FireCraft"), Some("44179611600")), + None + ); + assert_eq!( + built_run_url(Some("https://github.com"), None, Some("44179611600")), + None + ); + } + #[test] fn the_manifest_keys_and_entries_are_sorted() { let scratch = Scratch::new("sorted");