Skip to content

docs: first-class web (#143) audit + campaign prep (specs + roadmap) - #144

Draft
intendednull wants to merge 2 commits into
mainfrom
worktree-web-firstclass-audit
Draft

docs: first-class web (#143) audit + campaign prep (specs + roadmap)#144
intendednull wants to merge 2 commits into
mainfrom
worktree-web-firstclass-audit

Conversation

@intendednull

Copy link
Copy Markdown
Owner

Prep for the campaign addressing issue #143 ("Epic: Buiy first-class web"). Docs-only; no implementation. Execution is blocked on the Dooduel M1 merge — this is the research → spec → plan prep, fresh-context review-gated.

What's here

True residual (the actual campaign)

Not "build web support" — it's (A) close the reach-widening spec's one deferral (D3 Rgba16Float float-less fallback) + pay down the verification/operability debt it named but never scheduled (a11y CI guard + real-AT reach; IME test + real-device OSK; a11y inbound actions + live focus; web golden; runtime macOS-modifier; Msaa::Off), plus (B) MVU Subscription ingest and (C) URL router. Item 8 (OG/SSR) is out of scope (app backend).

Review gate

Fresh-context: design soundness (ship-with-minor-edits, no blockers — both MAJORs + minors applied) + factual-anchor verification (26/27 anchors confirmed, the one wrong line-cite corrected — grounded against current origin/main, not drifting like #143).

Not in this PR (gated)

The proposed #143 rewrite + split into #143a/b/c + close Item 8 is drafted for owner review, not yet posted to GitHub. Execution of any track waits on the Dooduel M1 merge.

🤖 Generated with Claude Code

intendednull and others added 2 commits July 11, 2026 20:25
…in/main

Issue #143 was authored against the stale #85 baseline (4010753) and never
reconciled against the Dooduel web waves W1-W5 (#94-#102) or the governing
browser-reach widening spec (2026-06-30), which already landed Items 1/3/4/5/9.

10-dimension multi-agent reconciliation against real origin/main 6e07954 +
adversarial verify on the a11y-sink and IME dimensions + orchestrator
spot-checks. True residual is narrow: close the reach-widening spec's one
deferral (D3 Rgba16Float float-less fallback) + verification/operability debt
it named but didn't schedule, plus two net-new subsystems (WebSocket-into-ECS
networking, URL router) outside any existing spec. Item 8 (OG/SSR) out of scope.

Adds the report + docs/README catalog entry.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Prep for the campaign addressing issue #143 (per the audit
2026-07-11-web-first-class-audit.md). Execution is blocked on the Dooduel M1
merge; these are design docs only.

- specs/2026-07-11-mvu-subscription-ingest-design.md — Track B. Reframes #143
  Item 6: the framework part is the MVU §8 keyed Subscription ingest seam (no
  new framework deps; spawn_local + crossbeam/async-channel already in-graph),
  the app owns the transport (ewebsock, as dooduel_core does). Resolves the
  audit's Item-6 contradiction without breaching the transport non-goal.
- specs/2026-07-11-buiy-url-router-design.md — Track C. A web-only History-
  provider seam (mirrors ClipboardProvider) driving the MVU nav model, behind a
  targeted foundation scope carve-out (URL-view-navigation only; persistence +
  transport stay out).
- plans/2026-07-11-web-first-class-campaign.md — roadmap: Track A hardening +
  verification debt (on the 2026-06-30 reach-widening spec) + B + C, the #143
  rewrite/split map, the Dooduel-merge block (per-track + schedule risk), and
  the product-decision / track-don't-build carry-overs.
- docs/README.md — catalog entries.

Fresh-context review-gated (design soundness + factual anchor verification:
26/27 anchors confirmed, the one line-cite corrected).

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant