Repository navigation
Remove the vendored makepad VM; the fork's octoscript branch is the source - #23
Merged
Merged
Conversation
…ource Since #22 octoscript-core and octoscript-capabilities take makepad-script by git from OctoSense-org/makepad at 1fdce273 (the head of the fork's octoscript branch, merge of fork PR #1 which carried Octoscript's VM hardening). The vendored copy under vendor/makepad is no longer built or depended on, so remove it and repoint everything that referenced it. - Cargo.toml: drop the workspace exclude; .gitignore: drop the vendor target. - CI: the makepad-script and makepad-regex suites and the Miri re-entrancy regression now run as `cargo test -p <pkg>`, resolved from Cargo.lock, so no fork checkout or second rev pin is needed. - UPSTREAM.md: rewritten as the git-dependency policy (rev in the two manifests, bumped together with OctoSense and Octoscript-Makepad; no local patches carried here). - docs/vm-hardening-history.md: the old vendor/makepad/PATCHES.md, kept as a history note behind a preface saying the changes were ported to the fork's octoscript branch on 2026-09-13. - CONTRIBUTING.md, README.md, NOTICE, docs/development.md, docs/positioning.md, docs/scoped-state.md, docs/makepad-ui-compatibility.md: describe the dependency instead of the import. Makepad and smallvec attribution in NOTICE is kept. Verified: cargo check --workspace --all-targets clean; cargo test --workspace --no-fail-fast 988 passed, 2 failed (pre-existing octoscript-ui-l0 profile.rs catalog-drift tests); RUSTFLAGS='--cfg fuzzing' cargo +nightly check --manifest-path fuzz/Cargo.toml clean. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01JgSir4cQyaakzPju4h6smw
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Since #22,
octoscript-coreandoctoscript-capabilitiestakemakepad-scriptby git fromOctoSense-org/makepadat1fdce273bf888239079754a1834734635eaed34f, which is the head of the fork'soctoscriptbranch (merge of fork PR #1, which carried Octoscript's VM hardening there). The vendored copy undervendor/makepadis no longer built or depended on, so this removes it and repoints everything that referenced it.Changes
vendor/makepadremoved (153 files).Cargo.tomldrops the workspaceexclude;.gitignoredropsvendor/makepad/target/..github/workflows/ci.yml): themakepad-scriptandmakepad-regexsuites and the Mirireentrant_reload_of_the_active_body_does_not_keep_an_opcode_pointerregression now run ascargo test -p <pkg>/cargo +nightly miri test -p makepad-script .... Cargo resolves the packages fromCargo.lock, so CI needs no fork checkout and no second rev pin. Verified locally that both-pselections build and the Miri test name resolves and passes.UPSTREAM.mdrewritten as the git-dependency policy: the VM comes from the fork'soctoscriptbranch; the rev lives in the two manifests (fuzz inherits it via path deps and carries no pin); bumping means moving the rev in both,cargo update -p makepad-script, running the VM suites, and moving OctoSense and Octoscript-Makepad to the same rev in the same window; local patches are no longer carried here.docs/vm-hardening-history.md(new): the content of the oldvendor/makepad/PATCHES.md, unchanged, under a preface stating the changes were ported toOctoSense-org/makepadoctoscripton 2026-09-13 (fork PR UI Profile Level 0: a canonical profile for LLM-authored UI #1) and no longer live here.CONTRIBUTING.md,README.md(baseline bullet, VM bootstrap paragraph, crate list),docs/development.md(VM test commands, parser-fix guidance, syntax-fuzzing note),docs/positioning.md,docs/scoped-state.md,docs/makepad-ui-compatibility.mdnow describe the dependency instead of the import.NOTICE: Makepad and smallvec attribution kept; only the "vendored" wording and thevendor/makepad/libs/smallvec/path were adjusted, since the git dependency still ships that code.fuzz/Cargo.tomlneeded no change: it has no path dependency on the vendor tree.Verification
cargo check --workspace --all-targets: clean.cargo test --workspace --no-fail-fast: 988 passed, 2 failed. The two failures are the pre-existingoctoscript-ui-l0profile.rscatalog-drift tests (the_field_vocabularies_match_the_toml_spec,the_source_catalog_matches_the_toml_spec), unrelated to this change.RUSTFLAGS='--cfg fuzzing' cargo +nightly check --manifest-path fuzz/Cargo.toml: clean (the fuzz targets need thefuzzingcfg, as CI already sets).Not touched: rustdoc comments and function names in
crates/octoscript-core/src/lib.rs(e.g.restrict_vendored_module_surface) and prose inSECURITY.md,docs/grammar.md,docs/language-profile.mdstill say "vendored VM/parser". They contain no paths and are semantically loose rather than wrong; renaming them is a separate cosmetic pass.🤖 Generated with Claude Code
https://claude.ai/code/session_01JgSir4cQyaakzPju4h6smw