Release 0.20.14: saved views, reliable forks and useful feedback - #190
Merged
Merged
Conversation
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Release 0.20.14 with the merged saved MCP view recovery (#186), reliable fork boundaries (#188), explicit internal-history filtering (#187), and actionable feedback diagnostics (#171). Full Python checks now run locally or on demand (#189); package/frontend/runtime qualification remains required for publication.
No CLI-based memory workaround is included. Internal-history discovery reads shared metadata directly and requires no CLI package or executable. Older unmarked jobs remain visible; the full canvas revision timeline and CLI-free portable memory inference are separate work.
Validation: 319 frontend tests passed, integrated recovery/fork/diagnostic regressions passed (75 passed, one existing opt-in runtime skip), 73 feedback tests and 172 history/navigation tests passed, release package verification passed. The full Python suite runs locally and its exact result will be recorded in the release handoff. Its initial macOS run exposed Linux service fixtures selecting the real local launchd backend; that run was stopped, the proven test-owned service was unloaded, and its files were retained as evidence. The fixtures now default to the mocked systemd contract, isolate both service paths and fail any unmocked service-manager call. All 65 service/restart/artifact-runtime regressions pass after this test-only correction. No Spark service was involved. Package and frontend versions, lock metadata and embedded release notes are aligned; generated assets are included.
Spark-1 remains on its normal published-release updater. No manual deployment, session inspection, history replay or active-chat interruption.