Release routes remain inspectable, but installer handoffs and availability or stable-release claims are withheld until immutable release authority is ready.
The project is moving from canon and architecture into a more coherent public product: clearer entry surfaces, hardened rules, stronger shared UI, and a more credible path toward release.
Support, feedback, and governance are separate lanes. Package/install claims stay tied to current release, journey, and rollback evidence instead of collapsing into one generic trust message.
Package, install, update, rollback, and revoke claims follow the current immutable release authority.
The public route only claims the flagship workbench when executable proof and veteran-familiar desktop signals agree.
Account value is framed as claim, restore, second-device continuation, and support continuity, not as a browser ritual.
Support is a closure lane tied to install and release receipts. It is not the same thing as votes, proposals, or roadmap wishes.
Feedback is an advisory intake lane. It informs governed discovery work, but it never masquerades as support closure or shipped rollout authority.
Governance cards describe launch, canary, rollback, and claim-safety posture. They do not imply immediate rollout just because a proposal exists.
These are the product parts a normal reader can understand without caring how the repos are split. Each line shows weighted progress, recent momentum, and a near-term ETA based on the current milestone map and the recent delivery pace.
Deterministic mechanics, reducer truth, and compatibility-safe runtime hardening.
Character building, inspectors, dense data UX, and the reusable Chummer design system.
Player and GM shell, local-first sync, browser session transport, and offline-resume play surfaces.
Accounts, groups, sponsorship, public entry, publishing, registry, and media execution services.
Product canon, horizons, participation workflow, public landing rules, and the public-facing guide bundle.
This is the public-facing version of the roadmap: one timeline per product part, focused on what changed, what is actively being closed, and what a complete checkpoint should look like.
Hardening wave. Legacy purification, retention controls, observability slices, and stronger verifier guardrails moved forward.
F1 closure. Finish observability, disaster recovery, and replay-safety proof.
Auditable boring runtime. A boring, auditable core runtime with no hidden legacy cargo.
Scope clarified. Remaining work was sliced into concrete queues for relationship graph, heat/favor, calendar, inspector, hub UX, portrait, and NPC persona.
Workbench complete. Close workbench-complete and quality-hardening waves; centralize reusable UI primitives.
Polished workbench. Complete desktop/browser workbench with accessibility, localization, and polished shared components.
Play-shell hardening. Play-shell hardening and browser transport/offline resume work was republished as runnable backlog.
E1 closure. Drive the final Play complete milestone across live-session depth and installable hardening.
Reliable session shell. A reliable session shell that survives weak networks and resumes cleanly.
Public entry wave. Public landing shell, email auth delivery, sponsor tiers/account linking, and canon refreshes across registry/media moved forward.
Completion wave. Finish adapter waves, hub-complete surfaces, docs/operator projection, assistant plane, and media-plane completion.
Public participation stack. A cohesive hosted community and publication stack ready for real public participation.
Canon tightened. Public landing/auth/download canon, participation workflow, sponsorship backlog, and release-governance wave were tightened.
Validator convergence. Finish release-completion proof and stronger validator-backed convergence across downstream consumers.
One public story. One authoritative public story, compiled cleanly into every downstream surface.
The public progress report, product-part grouping, and timeline view now read like one coherent status surface instead of scattered repo signals.
Public milestones, guide surfaces, and landing routes now line up more cleanly around user-visible product areas instead of internal execution seams.
Landing, auth, guide, and participation surfaces were tightened so the public front door stays closer to the actual product truth.
The remaining desktop/browser UX work has been sliced into concrete milestone- and backlog-shaped queues instead of sitting as fuzzy later work.
The current core wave is no longer about who owns what; it is about observability, disaster recovery, replay safety, retention, and finishing the last legacy purification passes.
The hosted stack is visibly moving, especially around public entry and email identity, but it still carries the largest remaining completion wave because registry depth, media execution, adapters, and broader hub product surfaces remain open.
Progress is a weighted estimate built from the current milestone map, grouped into public-facing product parts instead of repo names. The ETA is a momentum-based planning band, not a promise: it uses remaining milestone weight, recent repo activity, and uncovered scope breadth as a public proxy until weekly historical snapshots are published. ETA is a momentum-based planning band, not a delivery promise. Weekly public progress history is now available; ETA uses recorded velocity where enough snapshots exist and falls back to the short-horizon momentum proxy elsewhere. Mapped public milestones are complete, but the full-product frontier remains open with 3 milestone(s). Full completion requires full frontier closure. The current supervisor ETA is tactical or otherwise not yet calendar-grade enough for a full-product parity forecast, so the public surface tracks the remaining frontier without publishing a parity date band.
If you want to follow the rollout or join future contribution waves, the participation page explains the current public entry path. Open the participation page
The most interesting change is not a single feature. It is that architecture, execution, public entry, and product canon are starting to reinforce each other. That is the point where a serious internal system turns into something people can eventually use, trust, and follow.
