The trail ahead.

Set the focus on single-wheel trail riding, exploration, and a board without a rider character.

Established the controller and camera foundation used for subsequent riding experiments.

Compared multiple tiled-world sizes and narrowed the working island scale to roughly 2.56 km across.

Explored terrain streaming and biome workflows, with URP selected as the current direction for further performance work.

Iterated on route spacing, mountain coverage, shoreline and slope validation, and per-trail editing and bake requirements.

Worked through board proportions, materials, part visibility, and customization controls; developed the website model preview.

Build on the current controller: tune slopes, reverse lift, pushback, tight turns, and small obstacles through repeatable riding tests.

Finish keyboard/controller consistency, swivel and free-look behavior, and predictable return to the behind-board view.

Regression-test the streaming fall-through fixes, spawn scale, recovery distance, and unwanted reset behavior across the full island.

Build on the direct-connect prototype and settle the relay or NAT-traversal path. Separate basic connection testing from the final no-port-forwarding experience.

Check board movement, spawning, disconnects, and terrain loading when riders travel into different parts of the island.

Run small sessions over real internet connections; refine latency handling, reconnect behavior, and the session flow.

Build on the selected world scale and biome masks. Lock the main connections, landmarks, and route distribution instead of repeating the initial terrain experiments.

Improve island-wide spacing, mountain routes, shoreline checks, grade and corner validation. Finish the editable per-trail spline and targeted bake workflow without repainting unrelated terrain.

Continue creek carving and bank dressing, improve grass visibility and distant scenery, and profile terrain streaming, vegetation, and loading stalls in actual builds.

Preserve the supplied board proportions while checking face visibility, branding removal, grip, tire, hub, footpads, and bumpers. Verify each paint control affects the intended part.

Refine the existing customization and save work into a clear interface. Check straight rails by default, optional accessories, and compatibility for each board's available parts.

Verify saved setups after restarting and synchronize supported board/part/color choices when joining or reconnecting to multiplayer.

Connect trail names, descriptions, loop/point-to-point types, and readable start/end locations to the editable route data.

Integrate and test start/finish triggers, checkpoints, run guidance, results, restarts, and saved records; verify reversed or interrupted attempts are handled clearly.

Evaluate recording and replaying player-created start/end segments. Optional for the first release if validation or save work threatens the core schedule.

Finish the play flow, input and graphics settings, pause/recovery behavior, and a short introduction that does not require developer help.

Expand invites after the focused tests hold up. Check exploration, timed trails, garage persistence, and joining friends across a wider range of PCs.

Reserve time for reproducible crashes, confusing flows, controller issues, save failures, and changes that improve a complete session.

Expand testing gradually and check long sessions, varied hardware, separate-network multiplayer, and save compatibility between builds.

Finalize the initial feature scope, supported hardware, Steam store materials, build distribution, and update workflow. Prepare the store page earlier during alpha; allow time for platform review.

Target September within Q3 2027. Release only if the core test criteria hold; move the date if critical issues remain. Larger online features and shared player-created content can follow later.