展示HN:Waycore – 一款开源的、优先离线的模块化现场计算机
你好,HN,
我正在构建Waycore,这是一个开源项目,旨在探索灵活的、优先离线的户外计算机在生存和离网场景下应具备的特性。
项目的核心目标是适应性和韧性:
- 模块化硬件(外部传感器/工具模块)
- 可扩展的操作系统,支持外部应用(相关指导正在制定中)
- 无需互联网连接——地图、模型和知识可以离线使用
- 可选的LTE/Wi-Fi连接,需在可用时明确启用
项目的一个主要关注点是设备上的自主智能AI,而不仅仅是聊天或图像识别。该AI旨在:
- 读取实时传感器数据(GPS、指南针、环境)
- 基于离线知识进行推理
- 使用应用程序和核心API
- 协助导航、安全检查、记录和通信
主要项目代码库(操作系统与架构):
[https://github.com/dmitry-grechko/waycore](https://github.com/dmitry-grechko/waycore)
此外,还有一个独立的代码库,整理了可供离线使用的免费生存与户外PDF文档:
[https://github.com/dmitry-grechko/waycore-knowledge](https://github.com/dmitry-grechko/waycore-knowledge)
我希望能收到关于以下方面的反馈和贡献者:
- 针对坚固触控设备的UI/UX设计
- 硬件模块化及接口
- 离线/边缘智能体架构
- 在没有互联网的情况下表现良好的小型模型
- 高质量的公共领域或宽松许可的生存知识来源
欢迎提问或提出批评意见。
查看原文
Hi HN,<p>I’m building Waycore, an open-source project exploring what a flexible, offline-first field computer should look like for outdoor, survival, and off-grid scenarios.<p>The core goals are adaptability and resilience:<p>modular hardware (external sensor/tool modules)<p>extensible OS with support for external apps (guidelines in progress)<p>no required internet connection — maps, models, and knowledge work offline<p>optional LTE/Wi-Fi when available and explicitly enabled<p>A major focus is on-device agentic AI, not just chat or image recognition. The AI is intended to:<p>read live sensor data (GPS, compass, environment)<p>reason over offline knowledge<p>use apps and core APIs<p>assist with navigation, safety checks, logging, and communication<p>Main project repo (OS & architecture):
<a href="https://github.com/dmitry-grechko/waycore" rel="nofollow">https://github.com/dmitry-grechko/waycore</a><p>There’s also a separate repo curating freely downloadable survival & outdoor PDFs for offline use:
<a href="https://github.com/dmitry-grechko/waycore-knowledge" rel="nofollow">https://github.com/dmitry-grechko/waycore-knowledge</a><p>I’m looking for feedback and contributors around:<p>UI/UX for rugged touch devices<p>hardware modularity & interfaces<p>offline/edge agent architectures<p>small models that work well without internet<p>high-quality public-domain or permissive survival knowledge sources<p>Happy to answer questions or hear critique.