Working locally today
Connected systems that can be played and tested on a development PC. They remain rough, narrow and subject to revision.
In active development
Development journal
Built in visible, testable steps
Aethercell is an ambitious game made through deliberately small prototypes. This page separates what works today from what is being built and what still belongs to the wider design vision.
- Current phase
- Technical prototype
- Current access
- Local PC playtests
- Public release date
- Not announced
One project, three levels of certainty
Read every feature in context.
A functioning experiment is not a finished feature, and a design direction is not a promise. These labels appear throughout the site.
Being built or replaced
Current implementation priorities with real work behind them, but without a guaranteed public date or final form.
The direction of the game
Systems and experiences that guide present decisions. Their exact scope depends on what later prototypes prove.
Playable technical prototype
What can be tested now
A third-person Garden Lab with a persistent local state
The current build is not publicly downloadable. In local playtests, it connects camera-relative WASD movement, mouse camera, strafing, sprinting and zoom with autonomous activity, garden interactions and persistence.
Its value is not visual finish. It shows whether a player can enter a profile, move through the world, affect it, leave safely and return to a state the game can explain.
- Third-person movementCamera-relative WASD movement, continuous strafing and sprinting across a connected technical world.
- Camera and zoomA freely rotating mouse camera with controlled zoom for exploration and interaction.
- Shared action rulesPlayer and inhabitant actions use the same authoritative resolution path.
- Garden interactionsA bounded activity for testing action, time and world-state changes.
- Local identityA local profile with no online account or cloud dependency.
- Checkpoint persistenceValidated local saving, restoration and continuation.
- Seated logoutAn approximately 20-second in-world delay, followed by a final save, profile selection and restored continuation.
- Observation toolsDevelopment-only analytics and event views for understanding outcomes.
In active development
Current priorities
Make the next minute of play believable.
The present focus is a coherent, readable player experience—not a broad checklist of disconnected systems.
Character and animation
Replace the technical figure with a believable character, stronger locomotion and a visual scale that supports the intended world.
Enemies and close combat
Establish readable opponents, blocking, hits, damage and consequences before expanding combat variety.
The dual ring
Refine a two-ring action language for physical choices, resonant magic and context without turning play into menu management.
Controller support
Make movement, camera and action selection feel deliberate across keyboard, mouse and controller.
Feedback
Connect animation, sound, impact, light and interface response so an action is understood before it is explained.
Living context
Keep every visible addition compatible with persistence, observation and the longer-term simulation instead of staging a disposable demo.
Selected milestones
Foundations before spectacle
This is a curated development path rather than a release history. Each step records what it enabled and what it did not solve.
-
June 2026 · Runtime foundations
A world state that can be trusted
- Outcome
- Versioned state, deterministic random streams and replayable simulation boundaries.
- Learning
- Visible features are easier to change when their results can be reproduced and inspected.
- Open edge
- Reliable foundations alone do not create an emotionally legible game.
-
Early July 2026 · Action and observation
From intent to consequence
- Outcome
- Validated actions, transactional ticks and a separation between committed events and diagnostics.
- Learning
- The game must know what actually happened before it can communicate that result to a player.
- Open edge
- Developer-readable data still needs player-readable animation, sound and interface feedback.
-
Mid July 2026 · Persistent Garden Lab
A bounded world loop
- Outcome
- A local space joined movement, autonomous activity, garden actions, saving and restoration.
- Learning
- Small connected loops reveal more useful problems than isolated feature demonstrations.
- Open edge
- The tile-based laboratory established contracts, but not the intended camera, body or visual language.
-
Late July 2026 · Third-person technical prototype
Embodiment without pretending to be polished
- Outcome
- Camera-relative WASD movement, mouse camera, strafing, sprinting and zoom joined the persistent world loop, alongside a seated 20-second logout and restore path.
- Learning
- Camera, body, movement and persistence need to be tested together before visual polish can carry their meaning.
- Open edge
- The controls and continuity are technical foundations; character art, animation, combat feel, audio and accessibility are not finished systems.
What changed—and why
A prototype can succeed and still be replaced.
Aethercell treats discarded presentation as evidence, not embarrassment. Technical work remains useful when it clarifies the next decision.
The current third-person slice is a technical benchmark.
It proves connected movement, camera control, action, persistence and autonomous activity. Its technical character, animation and presentation are not the final visual target for the game.
Observation tools are not a player interface.
Development views help explain decisions and outcomes. Player-facing animation, sound and feedback still need their own deliberate language.
A technical demo is not a miniature finished game.
Temporary interfaces and narrow test spaces exist to answer questions. They should not be mistaken for promised content or final art direction.
Development method
Prove, observe, revise, connect.
The project advances through vertical slices that are narrow enough to understand and complete enough to expose interactions between systems.
Build a question
Choose a player-facing uncertainty, not merely a feature name.
Connect the path
Join input, rules, state, feedback and persistence in the smallest useful slice.
Test the contracts
Use automated checks for deterministic and persistent behavior, then perform human visual and play review.
Keep the learning
Preserve sound foundations, replace weak presentation and document what remains unresolved.
Follow the real project
The interesting part is not certainty. It is the next honest result.
Future updates will keep the same distinction between playable work, active development and long-term intent.