WC/DK — Principles of Animation Design for Knowledge Architecture
WorldCenterer Dynamic Knowledge — a Univrs foundational document
1. Thesis: gameplay = dynamic design
WC/DK treats a knowledge system as a place you move through, not a set of screens you look at. The founding claim: the interaction loop of a good puzzle game is a better model for working with knowledge than the form-and-dashboard loop of enterprise UI. A player reads a space, forms a hypothesis, acts, watches the world respond, and revises. A researcher, engineer, or analyst does the same with a body of knowledge. Current UI/UX hides that loop behind lists, tabs, and modals; WC/DK makes it the primary surface.
Each influence contributes one constraint:
| Influence | What it supplies | WC/DK constraint |
|---|---|---|
| ustwo, Monument Valley 1–3 | A proven grammar of readable, impossible, hand-crafted space; every frame a poster, every mechanic learned by touch | Legibility before cleverness. Motion must teach the space, never decorate it |
| Reinier de Graaf, Architecture Against Architecture (Verso, 2026) | A critique of the profession’s self-image: star authorship, waste, demolition, labor hidden behind the image | Build less, reuse more, show the labor. No new screen when an existing room can be re-read |
| Guy Debord, The Society of the Spectacle (1967) | The spectacle is a social relation mediated by images; images can replace lived relations | Anti-spectacle test. Any animation that is watched rather than played is suspect |
| BIG, Formgiving (TASCHEN, 2020) | Design as giving form to what has not yet taken shape; work sorted into six evolutionary strands | Form the future, not the feature. Motion vocabularies follow long trajectories, not sprint tickets |
The human premise behind all four: play is the condition of staying young, and sharing play is how a community communicates. In Huizinga’s terms, play happens inside a magic circle, a bounded space where ordinary rules are suspended and new ones hold (Calleja 2013 on Huizinga). A WC/DK world is a magic circle for knowledge: entering it should feel like opening a Monument Valley chapter; leaving it should feel like having understood something.
This is not decoration. Debord argued the Situationists’ job was preparing ludic possibilities to come, and that play must escape the competitive, spectator form (Debord, Contribution to a Situationist Definition of Play, 1958, via nettime). WC/DK’s playfulness is a political and epistemic choice, not a brand tone.
2. Influence 1: ustwo’s Monument Valley 1–3
Monument Valley is the closest existing thing to a WC/DK interface: a 3D world, read through a fixed isometric camera, where the player’s only verb is rotate, slide, or tap, and where the geometry itself is the puzzle. Three facts matter most for WC/DK.
It was born inside a UX studio, not a game studio. ustwo was a digital product design firm; the eight-person games team shipped the first game in April 2014 after roughly ten months of work, led by designer and artist Ken Wong, with Neil McFarland directing and Peter Pashley as technical lead (Game Developer, 2014). Wong’s GDC Europe 2014 talk was titled Less Game, More Experience (GDC Vault). The studio talks about being platform-perfect: the form follows the device it is played on (Metropolis).
Every screen must work as a poster. Wong’s stated goal was that each frame be worthy of public display (Wikipedia summary of sources). Lead producer John Lau restated it for MV3 in 2025: every level should be a piece of art you could print and hang, while still being a compelling puzzle (Unity, MV3: Blurring Art and Design). The levels are described as little worlds: microcosms surrounded by empty space (Kill Screen).
The illusion is a camera contract, not real geometry. The Penrose constructions only resolve at specific camera rotations; parts that are disconnected in 3D appear connected from the fixed view, and the character is allowed to walk across that screen-space seam (Kodeco tutorial; Sketchfab on illusion of continuity). Tooling: Unity, with Blender for modelling and animation (Creative Bloq).
How the three games evolved
| Game | Year | Director / lead | Core shift | Production lesson |
|---|---|---|---|---|
| Monument Valley | 2014 | Ken Wong (design, art), Neil McFarland | Rotate-and-connect on impossible architecture; one silent character, Ida | Levels blocked out from Unity cubes, then art applied by generating quads on visible faces. One-way handoff: design complete, then art |
| Monument Valley 2 | 2017 | ustwo games team | Two characters (Ro and child); emotional journey; split paths | In-engine cube modeler inspired by ProBuilder. Designer and artist edit the same GameObject, so the design–art loop closes |
| Monument Valley 3 | Dec 2024 (Netflix mobile); Jul 2025 PC and console | Jennifer Estaris (game director), John Lau (lead producer), Lili Ibrahim (lead artist) | Noor, a lighthouse keeper’s apprentice; sailing opens the space between structures; unfolding cubes shift planes | Decorating tools: artists build roof and facade meshes, then stamp them like brushes. The boundary between designing and arting dissolves |
Sources for the table: Unity, 2025; Engadget, 2024; Creative Review, 2025; Apple, Games We Love; ustwo games, 2025.
What WC/DK takes from it
- One camera, many truths. A fixed, rotatable view lets a knowledge space hold contradictions (two sources, two framings) that resolve only from a chosen angle. The seam is the insight.
- The verb set is tiny. Rotate, slide, tap. WC/DK should ship with no more than five spatial verbs; everything else is composed from them.
- The world responds, never the chrome. Staircases grow, towers rise, floors flip under the finger. There is no HUD telling you what happened. Feedback is architectural.
- Art and logic are the same object. The MV2 lesson (one GameObject edited by designer and artist) becomes a WC/DK rule: the data model and the rendered form are one artifact, so a change in knowledge is a visible change in architecture.
- Pacing has rests. MV3 added sailing as a meditative pause between dense puzzles (Creative Review). Knowledge work needs the same: drift segments between dense rooms.
- Platform-perfect, not platform-agnostic. The same world should rotate under a thumb, a mouse, a gaze-and-pinch, and a gamepad, with motion tuned per input, not averaged.
3. Influence 2: Reinier de Graaf, Architecture Against Architecture
The book is a 14-point manifesto, published by Verso on 24 March 2026 (272 pp., ISBN 9781804299036), by the OMA partner and AMO co-founder (Harvard Book Store; Blackwell’s). De Graaf calls it a workers’ manifesto: architecture should see itself as labor, not as the expression of a famous author (Dezeen interview, April 2026). It is split into two halves, Architects and Architecture, each essay anchored by an anecdote (ArchitectureNow review).
Positions reported across the reviews and interviews (the book itself was not read for this draft):
- Architecture is at war with the present; de Graaf calls for disarmament and conscientious objection (Bookshop.org blurb).
- Chapter two calls for unionisation; the manifesto prescribes retirement at 67, against the dominance of ageing star founders (Dezeen).
- Demolition is waste and a violent act toward those who live in and built a building (Verso page).
- The close: architecture should go beyond building and become a form of politics in its own right (Untapped Journal).
- Critics (Publishers Weekly, ArchitectureNow) say the diagnosis is sharp but the remedies and the vision of what next are thin.
Translation to WC/DK
De Graaf’s target is the image of the architect standing in for the work of building. Software UI has the identical disease: the launch video, the design-system showcase, the hero animation, while the knowledge underneath stays unreadable. WC/DK adopts five rules from the manifesto:
- No demolition. A knowledge room is never deleted when its data changes; it is renovated in place, and the renovation is visible (scaffolding, patched walls, dated layers). History is architecture.
- Authorship is plural. Every room shows who built it and who maintains it, in the room, not in a credits modal. The MV3 lesson (designer and artist editing one object) extends to contributors editing one world.
- Labor is legible. Rendering cost, data provenance, and maintenance debt are visible properties of a structure (weathering, load, light), not hidden metrics.
- Against the hero shot. No motion exists to be screenshotted. If a transition only makes sense in a trailer, it fails review.
- Platform as politics. WC/DK is built for Univrs as a publicly governed commons; the motion system is open and forkable, so no studio owns the vocabulary of how knowledge moves.
4. Influence 3: Guy Debord, The Society of the Spectacle
Debord’s 1967 book is 221 numbered theses. Its core definition: the spectacle is not a collection of images but a social relation among people, mediated by images; everything once directly lived has receded into representation (Wikipedia). The opening line détourns the opening of Marx’s Capital, replacing commodities with spectacles (EPFL concept page). Debord’s 1988 Comments added the integrated spectacle, a fusion of the concentrated (state) and diffuse (consumer) forms, which later scholars map onto networked media (Kellner and Best, Debord, Cybersituations, and the Interactive Spectacle).
Two points often missed: Debord’s project was positive as well as critical. The Situationists wanted constructed situations, dérive (drifting through a city by its psychogeographic pull), and détournement (re-using existing material against its original meaning) (nettime, Situationist gaming). And Debord opposed the competitive, spectator form of play, not play itself.
The anti-spectacle test
A 3D animated knowledge interface is at obvious risk of becoming a spectacle: a dashboard that is beautiful to watch and useless to act on. WC/DK applies four checks to every motion:
| Check | Question | Fails if |
|---|---|---|
| Agency | Did the user cause this motion, or is it playing at them? | Autoplay, ambient flourish, idle loops that carry no information |
| Relation | Does the motion show a relation between people or ideas, or only an image of one? | Animated counters, glow states, and hero transitions with no source behind them |
| Reversibility | Can the user drag it back and see the before state? | One-way reveals, cutscene transitions |
| Détournement | Can the user re-use this structure for a purpose the designer did not intend? | Locked layouts, non-forkable rooms |
Dérive becomes a first-class mode: drift, where the user moves through the knowledge world without a query, pulled by density, recency, or disagreement. The MV3 sailing segments are the formal precedent. Constructed situations become scenes: deliberately staged rooms that put two sources in tension so the user has to rotate the camera to reconcile them.
5. Influence 4: BIG, Formgiving: An Architectural Future History
Formgiving (TASCHEN, 2020; 736 pp., ISBN 9783836577045) is the third book in BIG’s TASCHEN trilogy after Yes is More (2009, a graphic novel) and Hot to Cold (2015). The title calques the Danish formgivning, to give form to what has not yet taken shape; the book is laid out as a logarithmic timeline from the Big Bang into the far future, and it was the companion to an exhibition at the Danish Architecture Center (TASCHEN; Metalocus). Note: the book is a TASCHEN title; O’Reilly Media is not a publisher of it.
The projects are grouped into six strands of evolution, which Ingels proposes as a map for projecting the future: Making, Sensing, Sustaining, Thinking, Healing, Moving (Metalocus; art4d). The book also includes Masterplanet, a crowdsourced master plan for the planet, and planetary proposals for Moon and Mars habitats.
Translation to WC/DK
Formgiving is the optimistic counterweight to de Graaf and Debord. Where they say stop and look at the labor and the image, Ingels says the future is unevenly distributed and architecture’s job is to give it form now. WC/DK uses the six strands as the top-level taxonomy of its motion vocabulary, so every animation primitive belongs to a trajectory, not a component:
| Strand | Knowledge-architecture meaning | Motion family |
|---|---|---|
| Making | Creating and editing nodes, rooms, structures | Extrude, carve, stamp (MV3 decorating brushes) |
| Sensing | Live data entering the world: telemetry, feeds, agent reports | Weather, light, tide; ambient but caused |
| Sustaining | Maintenance, provenance, debt, archival | Weathering, scaffolding, patina; renovation in place |
| Thinking | Reasoning paths, agent chains, argument structure | Staircases that connect, bridges that resolve only from one angle |
| Healing | Conflict resolution, reconciliation of sources, repair | Seam closing, floor flip, two views converging |
| Moving | Navigation, drift, hand-off between people and agents | Rotate, slide, sail, teleport with a visible path |
Two further rules from Formgiving:
- Timeline as home. The default view of any WC/DK world is a long time axis, logarithmic where needed, so a reader sees where a piece of knowledge sits between its origins and its projected consequences.
- Pragmatic utopia. Every speculative room (a projected policy, an unbuilt system) is rendered with the same material honesty as a built one, but flagged as projection in its lighting. The book’s claim is that dreams should already be tried pragmatically; the interface shows the trying.
6. The 3D Dynamic Animation System: what the canon says
There is no single product called the 3D Dynamic Animation System; the term covers three bodies of knowledge that WC/DK fuses. In 1987 John Lasseter (Pixar) published Principles of Traditional Animation Applied to 3D Computer Animation at SIGGRAPH, arguing that the Disney principles are essential to good 3D animation and should also matter to the designers of animation systems themselves (Lasseter 1987, PDF; SIGGRAPH record). Lasseter listed 11 principles; Thomas and Johnston’s The Illusion of Life (1981) is the source of the usual 12. The second body is motion dynamics: rigid-body and procedural simulation where mass, density, and material drive the motion instead of keyframes (Principles of Three-Dimensional Computer Animation, Norton 2003). The third is the modern platform motion systems: Material 3 Expressive (announced at Google I/O 2025) replaced fixed-duration easing with spring physics, because springs retarget from current velocity when interrupted while tweens snap (material_expressive package; compose-animation skill). Apple’s HIG motion guidance ties motion emphasis to input modality (stronger under direct touch, subdued under trackpad) and, for visionOS, warns that head-locked content and large moving objects filling the field of view cause discomfort (Apple HIG, Motion; WWDC23, Design considerations for vision and motion).
Lasseter’s principles, re-read for a knowledge world
| Principle (Lasseter 1987) | Character meaning | WC/DK knowledge-space meaning |
|---|---|---|
| Squash and stretch | Mass and rigidity shown by deformation | A node’s certainty is its rigidity: well-sourced facts barely deform under drag; contested claims wobble |
| Timing | Spacing of frames defines weight and personality | Spacing encodes cost: heavy structures (large corpora, slow queries) take longer and show it |
| Anticipation | Preparation before an action | A room telegraphs what a rotation will reveal before it completes; no surprise cuts |
| Staging | Present one idea clearly at a time | One room, one question. Camera and light suppress everything not in play |
| Follow-through and overlap | Parts keep moving after the main body stops | Dependent nodes settle after a parent moves; the settle shows the dependency graph |
| Straight-ahead vs pose-to-pose | Improvised vs keyed motion | Drift mode is straight-ahead; guided journeys are pose-to-pose with named poses |
| Slow in, slow out | Ease at the ends of a move | Springs, never linear; interruption is always allowed |
| Arcs | Natural motion follows curves | Navigation paths are arcs across the structure, never teleports, so the reader keeps orientation |
| Exaggeration | Push the essential quality | Disagreement between sources is exaggerated in geometry (a visible gap), not flagged in text |
| Secondary action | Supporting motion that reinforces the main one | Provenance trails, citation threads, and agent breadcrumbs move with the primary object |
| Appeal | Design the audience enjoys watching | Replaced by playability: a structure is appealing if a newcomer wants to touch it |
Two principles get special treatment. Solid drawing (from the 12) becomes solid geometry: every impossible seam is a deliberate camera contract, as in Monument Valley, never a rendering bug. Appeal is demoted below the anti-spectacle test in Section 4; a thing that is only pleasant to watch fails.
What the platform systems settle for us
- Springs are the default integrator. Spatial motion (position, scale, rotation) runs on stiffness, damping, and initial velocity; only effects (opacity, color) may use duration curves. This follows M3 Expressive and the Cupertino presets (Motor, unified motion API).
- Input modality scales emphasis. Touch gets the most physical response, pointer less, gaze-and-pinch the least and the slowest, per Apple’s HIG motion guidance.
- Reduced motion is a first-class mode, not a fallback: the same knowledge must be reachable with cuts and cross-fades, as the HIG and every modern motion skill insist (IxDF on UI animation accessibility).
- Comfort limits for immersive modes. Avoid head-locked content, oscillation, and large virtual objects crossing the field of view, per WWDC23 guidance. A WC/DK world on a headset must stay world-locked and let the user own the camera.
7. The WC/DK Foundational Principles
Twelve principles, one per line, each with the influence it descends from and the test that enforces it. A principle without a test is a slogan.
- Youth is the state of play. The system exists to keep its users curious. Test: a first-time user touches the world within 10 seconds without reading instructions. (Huizinga; ustwo)
- Every frame is a poster; every poster is a puzzle. Nothing is rendered that is not both beautiful at rest and answerable in motion. Test: screenshot any state; a stranger can say what question the room is asking. (ustwo)
- The world responds, the chrome does not. Feedback is architectural: structures grow, rotate, settle, weather. No toasts, badges, or spinners. Test: remove all 2D overlays; the state of the system is still readable. (ustwo; Debord)
- Five verbs. Rotate, slide, tap, drift, stamp. Everything else is composed. Test: any feature spec names which verbs it uses; a sixth verb needs a governance vote. (ustwo; de Graaf)
- Played, not watched. Motion that happens without the user’s cause carries live information or does not happen. Test: the anti-spectacle table in Section 4. (Debord)
- The seam is the insight. Contradiction is shown as impossible geometry that resolves from one angle; finding the angle is understanding. Test: every contested node has at least two camera positions with different truths. (ustwo; Lasseter’s solid drawing)
- No demolition. Knowledge is renovated in place; history stays as layers. Test: any room can be scrubbed back through its revisions as a continuous motion. (de Graaf)
- Labor is legible. Who built it, what it cost, and how it is maintained are material properties of the structure. Test: provenance is readable without opening a panel. (de Graaf; Debord)
- Springs, never tweens, for anything that moves. Spatial motion is physical and interruptible; the user can grab a transition mid-flight. Test: no spatial animation blocks input. (M3 Expressive; Apple HIG)
- Rests between rooms. Dense structures are separated by drift segments. Test: no journey chains more than three dense rooms without open water. (MV3 sailing; dérive)
- Form the trajectory, not the feature. Every motion primitive belongs to one of six strands: Making, Sensing, Sustaining, Thinking, Healing, Moving. Test: the motion token file has no orphan primitives. (BIG)
- Platform-perfect, forkable, and governed as a commons. Motion is tuned per input modality and per comfort setting; the vocabulary is open-source and owned by the Univrs community. Test: reduced-motion and headset modes ship in the same release as the default. (ustwo; de Graaf; Univrs charter)
The one-line version
A WC/DK world is a magic circle for knowledge, built from five verbs and six strands, where every frame is a poster, every seam is an insight, nothing is demolished, nothing merely plays at you, and the labor shows.
8. Knowledge architecture grammar
WC/DK maps the ordinary parts of a knowledge graph onto architectural parts with native motion. The mapping is the contract between the data model and the renderer: a change in one is a change in the other (the MV2 single-GameObject rule).
| Knowledge concept | Architectural form | Native motion | Strand |
|---|---|---|---|
| Node (fact, document, entity) | Block or room | Extrude on create; settle on edit; weather with age | Making, Sustaining |
| Edge (citation, dependency, causal link) | Walkway, staircase, bridge | Grows between rooms; collapses if a source is retracted | Thinking |
| Contested relation | Penrose seam | Connected from one camera angle, broken from others | Healing |
| Lens (filter, role, perspective) | Camera position and light | Rotate; a lens is a saved camera, never a hidden set of nodes | Moving |
| Journey (guided path, tutorial, argument) | Chapter | Pose-to-pose camera with named poses; rests between dense rooms | Moving, Thinking |
| Drift (unguided exploration) | Open water | Straight-ahead sailing pulled by density, recency, or disagreement | Moving |
| Live source (feed, telemetry, agent report) | Weather, tide, light | Ambient but caused; a storm is a spike with a visible origin | Sensing |
| Agent (MCP or A2A participant) | Lightkeeper, boat, crow | Visible actor with a path; never an invisible mutation of the world | Moving, Thinking |
| Revision history | Strata | Scrub back through layers in one continuous motion | Sustaining |
| Community (who maintains the room) | Scaffolding, inhabitants | Present in the room; contributors leave visible marks | Sustaining |
The rooms sit inside a World, the World sits on a Timeline (the Formgiving home view), and the whole thing is entered through a Threshold that marks the magic circle: a short, user-driven rotation that settles the camera and announces the rules of this world. Leaving is the same motion reversed.
Grammar diagram (text form)
FIVE VERBS (input) THRESHOLD (the magic circle) SIX STRANDS (Formgiving)
┌──────────────────────────────────┐
Rotate │ Timeline (home view) │ Making
Slide ── act on ─►│ ┌────────────────────────────┐ │ Sensing
Tap │ │ World │ │ Sustaining
Drift │ │ [Room] [Walkway] [Seam*] ├──┼── named by ─► Thinking
Stamp │ │ Open water: drift between │ │ Healing
│ │ Strata: history, never │ │ Moving
│ │ demolished │ │
│ │ Agents are visible actors │ │
│ └────────────────────────────┘ │
└──────────────────────────────────┘
* Seam = accented element: where understanding happens
Input verbs enter from the left and act on the nested forms inside the threshold; every resulting motion is classified by one of the six Formgiving strands on the right. The seam is the accented element because it is where understanding happens.
9. Technical blueprint
The blueprint follows ustwo’s own stack and the platform motion systems, adapted to a web-first, forkable commons.
Engine and rendering
| Layer | Choice | Reason |
|---|---|---|
| Scene engine | Three.js or Babylon.js in the browser; Unity for native builds when the web path is insufficient | ustwo shipped all three games in Unity, but WC/DK must run inside a browser and inside existing platforms (Barnyard, Imaginarium) without an install |
| Camera | Orthographic, fixed elevation (about 35°), rotation quantised to 90° stops with spring settle | Monument Valley’s illusions only resolve at exact camera angles; quantised stops make seams reliable |
| Impossible geometry | Explicit seam objects: pairs of disconnected faces registered as connected at one camera stop, with a walkable graph edge created only at that stop | This is the Kodeco and Paradox Toolkit approach: illusion points are modelled, not hoped for |
| World model | One object per room carrying both data and form (the MV2 single-GameObject rule); CRDT-backed so contributors edit the same object concurrently | Matches Univrs’s existing CRDT infrastructure and the no-demolition rule |
| Level authoring | In-engine cube modeller plus decorating brushes (roofs, facades, strata), as ustwo built for MV2 and MV3 | Lets the same person design and art a room, collapsing the handoff |
Motion tokens
All spatial motion is spring-based; effects use short duration curves. Token names follow the six strands.
token,family,stiffness,damping,initial_velocity,notes
moving.rotate.touch,spring,170,20,from_gesture,90-degree stop then settle
moving.rotate.pointer,spring,170,24,0,less overshoot than touch
moving.rotate.gaze,spring,120,28,0,slowest and most damped
moving.drift.sail,spring,40,12,from_gesture,low stiffness for open water
making.extrude,spring,220,18,0,visible overshoot on create
thinking.walkway.grow,spring,150,22,0,staircase or bridge grows end to end
healing.seam.resolve,spring,200,26,0,snap only at camera stop
sustaining.strata.scrub,spring,90,30,from_gesture,continuous history scrub
sensing.weather,curve,,,,duration 600 to 2400 ms caused by a live event
reduced.any,cut,,,,cross-fade 120 ms or hard cut
Numbers are starting points for tuning; stiffness and damping are in the dimensionless conventions used by M3 Expressive and Cupertino spring presets.
Input modality and comfort
- Emphasis scales by input: touch > pointer > gamepad > gaze-and-pinch, per Apple’s HIG motion guidance.
- Headset mode: world-locked only, no head-locked UI, no oscillating motion, large structures never cross the field of view uninvited (WWDC23 guidance).
- Reduced motion: same information, cuts and cross-fades, shipped in the same release as default motion.
- Interrupt everywhere: any spring can be grabbed mid-flight and retargeted from current velocity.
Performance budgets
| Target | Budget |
|---|---|
| Frame time, mobile web | 16.6 ms (60 fps) for a world of up to 200 visible rooms |
| Frame time, headset | 11.1 ms (90 fps) minimum, 8.3 ms preferred |
| Time to first touch | under 2 s from URL to a rotatable world |
| Draw calls per world | under 300 by instancing rooms and strata |
| Poster test | any frame exports at print resolution without a special render path |
Governance and openness
The motion token file, the seam format, and the room schema are published under an open licence through Univrs, with a review rule: a new verb or a new strand needs a community vote, a new token under an existing strand needs a pull request.
10. Open questions and next steps
- Read Architecture Against Architecture in full and replace the review-sourced summary in Section 3 with chapter-level references
- Decide the camera elevation and rotation stops by prototyping one room with a Penrose seam in Three.js
- Pick the first real corpus (Sepahsalar essays, Imaginarium projects, or a Barnyard agent registry) and build its threshold, timeline, and three rooms
- Tune the spring tokens on a phone, a laptop, and a headset; record which values survive all three
- Run the anti-spectacle test on an existing dashboard the team dislikes and document what fails
- Write the governance rule for adding a sixth verb
Open questions: Should agents (MCP/A2A participants) be rendered as characters like Noor and the crow, or as weather? How does a user author a seam without a 3D tool? What does a WC/DK world look like when printed, since the poster test demands an answer?
Sources
Pages opened for this draft:
- Unity: Monument Valley 3, Blurring Art and Design (John Lau, May 2025)
- Game Developer: Designing the surprise mobile game hit Monument Valley (2014)
- GDC Vault: Designing Monument Valley, Less Game, More Experience
- Creative Bloq: The making of Monument Valley
- Kill Screen: Escher’s little worlds
- Metropolis: The makers of Monument Valley
- Engadget: Monument Valley 3 adds a sailboat (2024)
- Creative Review: Monument Valley 3 game (2025)
- Apple, Games We Love: Monument Valley 3
- ustwo games: Monument Valley 3 out now for PC and console
- Kodeco: How to make a game like Monument Valley
- Sketchfab: Impossible Penrose Snake
- Dezeen: Reinier de Graaf interview (April 2026)
- Verso: Architecture Against Architecture
- ArchitectureNow: book review
- Untapped Journal: How to turn architecture’s devolution around
- Harvard Book Store listing
- Wikipedia: The Society of the Spectacle
- Kellner and Best: Debord, Cybersituations, and the Interactive Spectacle
- nettime: Dissolving the Magic Circle of Play, Lessons from Situationist Gaming
- Calleja: Ludic Identities and the Magic Circle (2013)
- TASCHEN: BIG. Formgiving
- Metalocus: Formgiving, the new book by BIG
- art4d: BIG. Formgiving
- Lasseter 1987: Principles of Traditional Animation Applied to 3D Computer Animation (PDF)
- SIGGRAPH history record for Lasseter 1987
- Apple HIG: Motion
- WWDC23: Design considerations for vision and motion
- Motor: unified motion API with Material and Cupertino spring tokens
- material_expressive package (M3 Expressive, I/O 2025)
The Univrs Ecosystem
Research: metalearn.org · ardeshir.io · sepahsalar.org · univrs.io
Creative: imagine.univrs.io
Code: github.com/univrs · github.com/ardeshir