feat(renderer): SDF disk cache (closes #22) + npm packaging#50
Merged
Conversation
Cold launches re-baked every per-mesh 32³ R32Float SDF from scratch even though the same content always produces the same voxel data. Sponza's first 9 frames spent ~8 bakes/frame on this path; second launch spent another 9. This change content-hashes (positions + indices) at GPU upload time, checks a platform-appropriate cache directory, and `queue.write_texture`s the cached voxel bytes directly when the file exists — bypassing the GPU dispatch entirely. Misses fall through to the existing bake; the renderer encodes a copy_texture_to_buffer alongside each dispatch and persists the readback to disk after the frame's main submit. The next launch hits and skips the bake. The cache is best-effort throughout: a corrupt entry, missing dir, or write failure silently re-bakes. wasm32 has no filesystem path so load returns None and store is gated out — web builds bake every launch as before. Cache layout: - macOS / iOS / tvOS / watchOS: ~/Library/Caches/bloom/sdf - Linux / Android: $XDG_CACHE_HOME/bloom/sdf - Windows: %LOCALAPPDATA%\bloom\cache\sdf - 16 B header (magic + version + voxel_res) + 128 KB R32Float payload Sponza disk footprint: 68 × 128 KB = 8.7 MB (matches the issue's budget). Disk reads happen synchronously at upload — a 128 KB read from local cache is sub-millisecond. The synchronous device.poll(Wait) on flush blocks for the bake submission to finish before persisting; this is a cold-launch-only stall (~9 frames) and the bake itself is the bottleneck on those frames anyway. Async pipelining is a follow-up if the cold-launch stall ever shows up in profiles. 8 new unit tests cover hash stability, change-detection on positions and indices, count-vs-value distinguishability, store/load round-trip, miss handling, size validation, and bad-magic rejection. cargo test 74/0 (was 66/0) on macOS, wasm32 cargo check clean.
Rename from bloom to @bloomengine/engine (Perry FFI module ref updated to match) and add a files: allowlist so the tarball ships just TS sources, Rust crates, shaders/assets, the bloom_jolt shim, and JoltPhysics/Jolt — vendored at publish time rather than fetched at install so installs stay self-contained and reproducible. .npmignore is belt-and-suspenders against target/, pkg/, build/, and the multi-MB Jolt extras (Samples/Docs/UnitTests/Assets/…). scripts/prepack.sh refuses to publish if the Jolt submodule isn't initialised — deliberately not auto-initing so we don't silently publish stale refs. Root MIT LICENSE added so the legal terms travel with the package.
Now that the package is published to npm, swap every documented import from "bloom" to "@bloomengine/engine" so the snippets actually resolve against an installed package. README also gains an Install section up front pointing at npm (plus bun/pnpm/yarn equivalents) and the toolchain prereqs (Perry + Rust, wasm-pack for web).
Adds a publish-npm job to the existing tag-driven release workflow. Runs after github-release so a failed publish doesn't leave a release-but-no-package state, and after the await-tests gate so we never ship a tag that didn't pass CI. Idempotent: re-checks npm before publishing and skips cleanly if the version already exists (so workflow_dispatch on an old tag won't double-publish). Checks out submodules recursively because the prepack hook refuses to ship without JoltPhysics sources on disk, and uses --provenance for the npm attestation badge. Requires an NPM_TOKEN repo secret with publish rights on the @bloomengine scope.
The package is now configured on npmjs.com with this workflow as a trusted publisher, so `id-token: write` is sufficient — npm publish exchanges the GitHub OIDC token for a short-lived credential. No long-lived NPM_TOKEN secret to rotate or leak, and provenance attestation is automatic.
proggeramlug
changed the title
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Two threads of work landed on this branch:
SDF disk cache (
eadf596, closes #22)On-disk cache for per-mesh SDFs, so cold restarts skip re-baking.
npm packaging (5 follow-up commits)
Prep the engine for
npm install @bloomengine/engine:JoltPhysics sources are vendored into the tarball at publish time (via `prepack` script enforcing submodule init) rather than fetched at install — keeps installs self-contained, reproducible, and free of `postinstall` script risk. Tarball is ~2.0 MB packed / 8.1 MB unpacked.
Test plan