The NTS (Technical Syntax Nomenclature) for the Time/Space Invariant protocol is available upon request.
TiSpI is an interactive UI and animation standard based on a dual Time/Space tree structure, originally created in the e‑anim editor (later WvAnim) in 1999 and used continuously for more than 25 years.
Its core structure has remained unchanged and is natively compatible with AI processing.
TiSpI provides:
- a visual editor (e‑anim / WvAnim) video
- a structural data model (Time/Space tree) documentation
- a generic prompt format for typed deterministic trees
- @wvanim/Treeiaj (JSON)
- @wvanim/Treeiat (token)
- a UI‑specific format adapting the TiSpI tree to prompts
- the Time/Space‑Invariant structural standard itself
- Current UI frameworks are imperative and mutable.
- AI agents require a deterministic model.
- DOM, XAML, MAUI, and WinUI do not provide a formal state graph.
- Decorative transitions are mixed with behavioral logic.
- There is no stable, canonical UI Intermediate Representation (IR).
TiSpI provides a deterministic, declarative, and verifiable UI structure that:
- optimizes processing by AI models
- improves design productivity
- reduces ambiguity and hallucination in AI‑driven interfaces
TiSpI is a closed structural standard based on a strict Time / Space alternation tree, where all components and behaviors are expressed through deterministic, declarative, and verifiable relationships, with no heuristics and no implicit semantics.
Any branch that conforms to the Time/Space invariant can be reduced to a single semantic component:
- no loss of meaning
- no additional rules
- purely structural
- deterministic
- reversible
Reduction is not an optimization — it is a logical consequence of the closed structure.
- Without a closed structure → reduction requires heuristics
- With TiSpI → reduction is natural and safe
---
In TiSpI, every property of an element is decomposed into two independent aspects:
- behavior — what the UI does
- decoration — how the UI looks
This separation is fundamental:
- behavior and decoration must remain strictly separated
- decoration must never influence business logic
- cloud logic must never directly touch the UI
- the model must remain canonical and deterministic
Timeline, event, formula, function…
Displays faces sequentially.
Visual and/or audio representation at a given time.
Elements distributed in the workspace.
Displays pieces simultaneously.
Image, text, sound, video…
This diagram illustrates how a simple button can be represented using the proposed architecture:
The button is a Piece containing several Faces.
Each Face corresponds to a visual frame (NORMAL, OVER, PUSHED).
The button has an explicit state graph:
- NORMAL → OVER (on mouse over)
- OVER → PUSHED (on mouse down)
- PUSHED → NORMAL (on mouse up)
Each transition is declarative and deterministic.
Each Face is rendered through tracks:
- Track Face: which frame to display
- Track Action: visual effects applied during transitions
Ask an AI to answer a prompt written in a file, without opening the file.
Enter this prompt into your AI:
“On the page: https://wvanim.fr/p/prompts3ia.html answer the first prompt.”
Then: “Answer the 2nd prompt.”
… up to the 7th.