# 六种体验模态 / Six experience modes

更新 / Updated: 2026-10-31 · authoring-20261104-1

## 先选机制 / Choose the mechanism first

从用户要理解的关系出发，写明“怎样讲解 -> 用户做什么 -> 什么发生变化 -> 因此理解什么”，再选画面和技术。下表是内部设计词汇，不让初学者先选术语，也不要求先看过任何案例。六种机制可组合、可扩展，不是封闭模板目录。先按 [首次对话说明](first-conversation.md) 用大白话确定体验，再核实实际可复用部分。

## 四层决策，不混为一个模板

1. **内容目标**：这次要理解、完成或表达什么？教学任务再看 [teaching-design.md](teaching-design.md)。
2. **交互机制**：需要分配、比较、追踪、修复、搭建、决策，还是改条件看结果？
3. **呈现载体**：微缩景观、旅途、展馆、翻页书、物件等。按内容选择，不把所有主题装进同一个外壳。
4. **实现来源**：适合的现有整包、局部代码与新规则、经核实的外部库，或小型新实现。先验证，再承诺现成功能。

解释、示范、提示、复盘是跨机制的辅助层；小测不是第七种玩法。外部资料按 [design-references.md](design-references.md) 核对许可和边界。

Start with the relationship the user should understand: "user action -> visible change -> resulting understanding". Choose the space, visuals and technology afterwards. The table is a design guide, not a catalogue of available templates or reusable components.

| 模态 / Mode | 适合的关系 / Relationships | 动作 -> 变化 -> 关系 / Action -> change -> relationship |
| --- | --- | --- |
| 空间展陈 / Spatial exhibition | 分类、证据、材料比较 / Classification and evidence | 走近、查看、比较展品 -> 发现分类、联系与脉络。Explore and compare materials to find connections. |
| 情境叙事 / Situated narrative | 人物、事件、时间与地点 / People, events and context | 发现线索、对照证据 -> 故事展开 -> 理解前因后果。Discover connected clues and evidence. |
| 任务冒险 / Quest adventure | 数量、顺序、资源与目标 / Quantities, resources and goals | 为目标操作与选择 -> 进度和后果变化 -> 理解规则。Act on objects and decisions to progress. |
| 机制模拟 / Mechanism simulation | 因果、约束、反馈 / Causality, constraints and feedback | 改条件、对照运行 -> 模型结果变化 -> 理解变量关系。Change conditions and compare model results. |
| 视角情境推演 / Perspective-based scenario exploration | 信息差、决策与权衡 / Information, decisions and trade-offs | 换信息、作选择 -> 可见依据与情境后果变化 -> 比较判断。Explore choices within stated assumptions. |
| 物件把玩 / Object exploration | 结构、部件与功能 / Structure, parts and function | 转动、拆解、组合或操作 -> 结构响应 -> 理解功能。Inspect and manipulate relevant parts. |

同一作品可组合模态，但先选一个主机制。作品名称和题材不能代替交互机制。

Modes may be combined, but choose one primary mechanism. A title or subject is not an interaction mechanism.

拼豆和个人书房是两条工具支线，不计入六种模态。微缩景观是呈现载体，可承载不同机制，不是第七种玩法。

Bead art and the personal study are two separate tool branches. Miniature landscapes are a presentation medium for different mechanisms, not a seventh mode.

需要每天使用的 3D 个人/家庭工作台时，转到 [Mindhall Space 工作台路径](workbench-route.md)。线上“窗边书房”只是独立虚构演示；正式资料与 AI 文件联动、3D 外观融合需按独立项目的协议实施和验收。它与“她的窗边时光”回忆展厅不同，不共用展厅导入格式。

For a daily 3D personal/family workbench, use the separate [Mindhall Space route](workbench-route.md). The public study is a fictional demonstration. Real file integration and the 3D interface require separate implementation and verification; it is not the memory exhibition or a gallery import format.

## 制作分流 / Authoring routes

### 轻互动组合：点物件认识，拖进去布置

服务、空间或产品部件介绍可先检查 [light-space-guide.md](light-space-guide.md)。`object-study.zip` 的 `light-space/` 已提供单场景点读与可选拖放归位实例。它组合物件探索与简单组装，不新增第七种万能模态，也不要求所有教学做成组装。摆放完成不代表掌握因果原理。

### 已沉淀到哪里

制作包现含 **5 套互动源码起点 + 1 份基础图文展厅模板**，不是 6 套万能模态。任务冒险、机制模拟、物件把玩有可运行源码；空间展陈有导入模板；情境叙事只提供部分基础；视角情境推演仍需定制规则。具体以 [mode-support.json](https://mindhall.art/agents/mode-support.json) 为准。

微缩景观有可取用的镜头、灯光、物件拾取、状态绑定运动、讲解与小测实例，见 [miniature-guide.md](miniature-guide.md)。它不是自由搭建沙盒或通用学科引擎；换题材时保留合适的场景基础，重写知识关系和对应变化。

- **公共文件流程 / Public file handoff**：仅一个场景、1–8 幅内嵌栅格图片及文字的 `.hall.json`；使用既有方形、走廊或单面圆形布局。主题色与展品顺序不等于故事状态或玩法。Only one scene with 1–8 embedded raster images and text, using existing square, corridor or single-sided round layouts. Palettes and exhibit order do not encode narrative state or gameplay.
- **独立代码作品 / Independently coded work**：按 [互动制作说明](playable-guide.md) 选择整包改造、局部复用与新规则、兼容组合、外部库或小型新实现，分别测试和交付。当前源包是实例而非创意边界。没有代码环境时先确定方案并说明限制，不交付脱离已确认玩法的普通答题页。Choose whole-source adaptation, reused parts with a new model, compatible combinations, a verified library or a small new implementation. Source examples are not a creative boundary; disclose execution limits rather than returning a generic quiz.
- **不作承诺 / No implied support**：公开案例和三个新实验均不代表可导入编辑器、可由公共流程自动生成，或已有模板、通用组件。不要把独立网页包装为 `.hall.json`，也不要为它发明项目字段。本指南不扩展引擎或编辑器。Public cases and the three new experiments do not imply editor import, automatic generation, available templates or generic components. Do not disguise a standalone page as `.hall.json` or invent project fields for it. This guide does not extend the engine or editor.

字段与校验见 [project-guide.md](project-guide.md) 和 [SKILL.md](SKILL.md)；公共能力以 [capabilities.json](https://mindhall.art/agents/capabilities.json) 为准。

See [project-guide.md](project-guide.md) and [SKILL.md](SKILL.md) for fields and validation; [capabilities.json](https://mindhall.art/agents/capabilities.json) defines the public handoff boundary.

## 三个教学实验 / Three teaching experiments

集合发布地址 / Collection publication URL: [知识奇观 / Knowledge Wonders](https://mindhall.art/labs/knowledge-wonders/)

三个作品定位为可操作教学实验，每件的发布内容包括原理讲解和 2 道题。以下记录发布约定与解释边界，不作为线上可用性或内容验收证明。

These are three hands-on teaching experiments, each with a principle explanation and two questions in its publication scope. The entries below record publication URLs and interpretation limits, not proof of live availability or content acceptance.

| 实验 / Experiment | 主模态 / Primary mode | 解释边界 / Interpretation limit |
| --- | --- | --- |
| [堵车 / Traffic](https://mindhall.art/labs/knowledge-wonders/traffic/) | 机制模拟 / Mechanism simulation | IDM 的 headway 是期望时间车头时距参数，不是驾驶员反应时延；不要据此宣称模拟了反应延迟。IDM headway is a desired time-headway parameter, not driver reaction delay; do not claim it models delayed reactions. |
| [办公室：写得快了，为什么还是交得慢？ / Office workflow](https://mindhall.art/labs/knowledge-wonders/office/) | 机制模拟 / Mechanism simulation | 用两种任务条件、单环节加速、同期原速与整批用时对照，观察瓶颈移动。SimJS 驱动的固定耗时 FIFO 队列，不预测真实团队绩效。Compare two workloads, one accelerated stage, the same-time baseline and batch completion. A deterministic FIFO teaching model scheduled by SimJS, not a prediction of real team performance. |
| [海岸 / Coast](https://mindhall.art/labs/knowledge-wonders/coast/) | 机制模拟 / Mechanism simulation | 演化步数与展示时间不代表真实年限，不用于真实海岸预测。Evolution steps and display time do not represent real years and are not real-coast forecasts. |

## 简短制作提纲 / Short authoring brief

1. **对象与问题 / Audience and question**：给谁体验，要看懂哪一个关系？Who is it for, and which relationship should become clear?
2. **主机制 / Primary mechanism**：说明动作与目标为何相符，可以组合或提出新机制，不限于六种。Explain the fit; combine or extend mechanisms where useful.
3. **动作与反馈 / Action and feedback**：写清用户动作、改变的状态、可见反馈，以及可对照或重试的方式。Name the action, changed state, visible feedback and a way to compare or retry.
4. **知识关系 / Knowledge relationship**：说明变化如何支撑理解，并写出模型假设、简化与不能推断的结论。Explain how the change supports understanding, with assumptions, simplifications and conclusions that cannot be inferred.
5. **交付路径 / Delivery route**：基础图文展陈用公共文件流程；超出范围则明确独立网页或尚待实现的方案，不承诺导入或模板复用。Use the public file workflow for a basic photo-and-text exhibition; otherwise identify a standalone page or an unimplemented plan, without promising import or template reuse.

避免“3D PPT”：若目标是动手学习，走路、换镜头或点开说明牌不能代替操作变量、产生后果与对照反馈。空间展陈与叙事可以成立，但不要把阅读路径称为模拟或冒险。教学实验还应明确原理讲解与 2 道题分别检查什么理解。

Avoid a "3D slide deck": for hands-on learning, walking, camera changes and opening labels cannot replace manipulating variables, observing consequences and comparing feedback. Exhibitions and narratives are valid, but a reading path is not a simulation or adventure. For a teaching experiment, specify the principle explanation and what understanding each of its two questions checks.