Skip to content

jedit Real Echo Release Gate #515

Open

Description

@flyingrobots

Migrated from Method backlog

This issue was created from a legacy filesystem backlog card. GitHub Issues are now the live work tracker; repository docs remain Method evidence.

Source backlog: docs/method/backlog/v0.1.0/PLATFORM_jedit-real-echo-release-gate.md
Original lane: v0.1.0
Original legend: PLATFORM

Original backlog card

jedit Real Echo Release Gate

Status: active release blocker.

Depends on:

Echo v0.1.0 is blocked until jedit works on Echo from the sibling jedit
repository.

Why now

The in-repo external contract fixture proved the generic mechanics. It did not
prove application ergonomics, cross-repo generated artifact use, real WASM or
host packaging, or the product pressure of a serious external application.

jedit is the release witness because it is external, product-shaped, and strict
about keeping application nouns out of Echo core.

Required path

sibling application-owned contract
-> Wesley generated Echo runtime artifacts
-> Echo installs a generic generated package
-> application submits canonical intent
-> trusted Echo host stages work and authorizes scheduler opportunities
-> Echo scheduler emits tick receipts
-> application observes intent outcome
-> application queries bounded reading
-> retained receipt and reading evidence can be inspected
-> replay reproduces the result

Acceptance criteria

  • A documented jedit command runs against a clean Echo checkout or
    published Echo artifact.
  • The proof uses generated jedit/Wesley contract artifacts rather than a
    second hand-authored protocol shape.
  • product capabilities remain application-side nouns; Echo core and Echo
    package boundaries do not define or import them.
  • opaque read-basis tokens remain supporting app-safe tokens, not primary
    runtime coordinates.
  • sibling application code does not access scheduler control.
  • scheduler start, until_idle, and fault recovery remain trusted host
    authority.
  • at least one application mutation uses non-trivial vars.
  • at least one bounded reading returns product-shaped payload plus Echo
    reading evidence.
  • at least one non-happy path is covered: unsupported operation/query,
    lawful rejection, admission obstruction, residual reading, or missing
    retention.
  • retained receipt and reading evidence can be loaded or inspected by
    semantic coordinate.
  • replay reproduces the same outcome and reading.
  • Echo core contains no application product APIs.

Current blocker

The existing jedit opt-in real Echo WASM witness still reflects an older model:
it tries to carry scheduler control through app-facing dispatch. Current Echo
correctly rejects that as forbidden control intent. The release work is to
update the jedit witness around the app/host split, not to weaken Echo.

Non-goals

  • Do not build all of jedit.
  • Do not move application product nouns into Echo.
  • Do not grant application code tick authority.
  • Do not replace jedit's fake transport harness before the real witness is
    ready.
  • Do not require full Continuum transport/import.
  • Do not require full observer-rights governance.

Activity

  1. coderabbitai commented on Jun 1, 2026

    @coderabbitai
    Contributor
    🔗 Related PRs

    #326 - Echo contract hosting roadmap [merged]
    #329 - Stack Witness 0001 fixture-backed jedit walking skeleton [merged]
    #365 - feat(wesley): emit query observer host helpers [merged]
    #369 - docs: add Echo v0.1.0 release plan [merged]
    #380 - feat(core): expose retained evidence posture on readings [merged]


    📝 Issue Planner

    Check the box below or use the @coderabbitai plan command to generate an implementation plan and prompts that you can use with your favorite coding assistant.

    • Create Plan

    🧪 Issue enrichment is currently in open beta.

    You can configure auto-planning by selecting labels in the issue_enrichment configuration.

    To disable automatic issue enrichment, add the following to your .coderabbit.yaml:

    issue_enrichment:
      auto_enrich:
        enabled: false

    💬 Have feedback or questions? Drop into our discord!

  2. modified the milestones: , Echo 1.0 on Jun 20, 2026
  3. flyingrobots commented on Aug 18, 2026

    @flyingrobots
    OwnerAuthor

    Architecture supersession note: this issue predates the now-explicit
    Jim.edict active-observer ownership model. Do not use it to add Jedit nouns,
    verbs, rope intrinsics, planners, or callbacks to Echo, or to make a TypeScript
    frontend call application operations directly as the final composition.

    The active generic-runtime work is #684, Interpret compiler-produced bounded
    Edict graph programs
    . It is blocked on flyingrobots/jedit#296 producing a real
    compiler-generated package from Jim-owned ReplaceRange.edict. Jedit's schema
    and oracle are ABI/conformance evidence only, never executable semantic input.
    The final consumer cutover is flyingrobots/jedit#295 and #301: Echo delivers
    canonical events to installed Jim.edict, which requests bounded readings,
    derives Jim-owned operation intents, handles outcomes, and advances editor
    state.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    lane:releaseMethod lane release.legend:platformMethod legend platform.release:echo-1.0Echo 1.0 release convergence scopespecSpec/Design document

    Projects

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions