Skip to content

InkKit 0.0.11

Choose a tag to compare

View all tags
@aicayzer aicayzer released this 10 Oct 11:36
· 185 commits to main since this release
v0.0.11
a0ccde4
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Complete InkKit's consolidation wave for generic minimal and rich offline hosts. No consumer app adoption or hosted playground deployment is included.

Changes

  • Add optional code highlighting and diagram preview controls. Disabled previews retain editable authored source and explicit clipboard/print output.
  • Cache successful source serialisation by immutable document. Representative large/mixed editing medians roughly halved; unchanged snapshots fell below timer resolution.
  • Fix multiline formatted insertion, inherited marks and terminal breaks; reject unsupported multiline table insertion before mutating source, revision or history.
  • Correct comment visibility and clipboard API documentation.
  • Extend generic offline native integration and reproducible bundle/startup/editing measurements.

Default rendering remains enabled. Minimal and rich builds retain the same single-file offline payload. The measured consumer grew from 6,086,749 to 6,095,767 bytes (gzip 1,720,663 to 1,724,624 bytes); no byte reduction or general startup improvement is claimed. See performance methodology and results and public API.

Verification

PR #54 resolves all demonstrated independent review findings. Required package/type/build/format/playground checks passed with 587 regressions. Eight selected Chromium cases and nine WKWebView scenarios (238 steps) passed against the actual candidate, including printing, source/TXT reloads, cancellation and instance privacy. Print preferences and clipboard were restored and compared. Performance evidence contains 18 fresh-page samples per version across both configurations and three document shapes.

Hosted Linux/macOS smoke and package checks passed in PR CI. Publication and registry comparison details follow below.

Limits

Whole-document preservation validation still dominates very large edits. Fresh-page Chromium timings do not establish universal cold-start or WKWebView timing. Synthetic composition checks establish guards, not full IME interoperability. Hosts still own storage, permissions and native actions; consumer app adoption remains separate.

Published package

Tag-triggered validation and OIDC publication passed at reviewed merge a0ccde4. Main CI passed before tagging. Signed provenance is recorded. npm latest is 0.0.11.

Registry SHA-256: 4c48a3ad9b832ca6fb8ea1825b30048f860d2bbd035be63ebc84b1a4014b0db3. Registry SHA-512 integrity: sha512-Cpp5kLT2xx+ZCreFtVPHzEjncraJk1IKpPBaYXaXxZ7DI2MDIg4fIQQIEfHXRL2nTjfR6SHCHgQDY0OsZhjg7Q==. The registry archive is byte-identical to the actual hosted release artefact. It differs from the native candidate only in compression; decompressed tar SHA-256 a5bad922d88f7cfefaf8c33cde4623c7304b43c2535249e5cdc7323e70c77ff3 matches exactly. A fresh isolated exact-version consumer passes public declarations, CSS and offline bundling; its bundle SHA-256 7a7515e3ea6248fc0340bb1c6a2511e9b5ccbc226a3df805593b11f9f5eeda9b also matches the candidate consumer.

Attached evidence includes candidate/hosted/registry identities, native inputs/results/PDF, restoration checks, independent review dispositions, and raw 0.0.10/0.0.11 performance samples. Full Atlas logs and failed attempts remain at /Users/aicayzer/Services/inkkit-worktrees/0.0.11/_local/release/0.0.11/.

Assets 12
Loading