Skip to content

Releases

Framefetch Desktop v0.2.0 · 正式桌面版本

Choose a tag to compare

@StephenQiu30 StephenQiu30 released this 20 Oct 07:32

Framefetch Desktop v0.2.0 正式版,配套 Server/Web v0.3.0。页面直接同步自该 Server 发行提交,连接已有服务;客户端不运行媒体引擎或 AI 后端。

本次正式版

  • 同步统一媒体校验后的错误反馈:终态失败显示手动重新获取/重新解析,仅等待重试时提示自动继续。
  • 汇集预览版后的共享页面、公众号来源选择、本地视频上传、内置视频/剧本 Skill 与报告入口、下载恢复及历史管理改进。
  • 保持 API 同源会话、WebSocket、HEAD/Range 文件传输、受限上传目标、外链处理和沙箱边界。

安装包

  • Framefetch-0.2.0-mac-arm64.dmg:macOS 14.0+,Apple Silicon。
  • Framefetch-0.2.0-win-x64.exe:Windows x64。
  • SHA256SUMS.txt:两个安装包的 SHA-256 校验清单。

安装包直接取自本发行提交的成功 CI 产物,不复用旧 Beta 安装包。包内版本为 0.2.0。

安装包未签名,macOS 未公证。 正式发行渠道不代表已完成签名、公证、商店审核或所有目标系统的干净安装/升级验收。当前没有 Intel macOS 或 Linux 安装包。

连接与验证

先部署 Server/Web v0.3.0,再用 Server 账户登录。默认本机服务为 http://127.0.0.1:8111/;远端使用 HTTPS 根地址。既有用户数据不自动删除或导入。

源码同步检查、lint/类型、42 项单元测试、构建及开发态传输 E2E 已通过。本发行还由 macOS ARM64 与 Windows x64 CI 各自打包并执行打包态传输 E2E。E2E 使用受控夹具,不能替代真实 Server 全业务、真实平台、干净安装、升级与卸载验收。截图视频四档完整文件交付的真实证据属于 Server/Web 验收,不冒充原生桌面验收。

中文说明 · English README

English summary

Stable desktop release paired with Server/Web v0.3.0, including updated media-validation recovery messages and the current shared workspace. Fresh macOS ARM64 and Windows x64 installers are built from the release commit; verify them against SHA256SUMS.txt.

Installers are unsigned and macOS is not notarized. Automated source/package transport E2E use controlled fixtures and do not establish full real-Server or target-system installation acceptance. A reachable Server is required.

完整变更 / Full changelog

发行证据 / Release evidence

发行提交 / Source commit · 本提交全部 CI / Exact-commit CI

FrameFetch Desktop v0.2.0-beta.1 · 首个桌面预览 / First desktop preview

Choose a tag to compare

@StephenQiu30 StephenQiu30 released this 03 Oct 05:13

把帧取 FrameFetch 工作区带到桌面:通过内置界面连接自己的 Server,在 macOS 和 Windows 上处理视频素材、阅读剧本并查看 AI 分析报告。

Bring your FrameFetch workspace to the desktop. Connect to your own Server through a bundled interface to work with video assets, screenplays and AI reports on macOS and Windows.

首个桌面预览版 / First desktop preview

  • 内置 React 页面、亮暗主题、字体与品牌资源,无需单独启动 Frontend 进程。

  • 共享 Server 的账户、权限、素材记录、剧本、任务与报告。

  • 支持本机服务和远端 HTTPS 部署,保存连接偏好,按 Server 地址隔离桌面会话。

  • 上传、文件下载、Cookie 与 WebSocket 沿用服务端协议。

  • 提供 macOS Apple Silicon DMG 与 Windows x64 安装包,以及 SHA-256 校验清单。

  • Bundled React pages, light/dark themes, fonts and brand assets, with no separate Frontend process.

  • Shared Server accounts, permissions, media records, screenplays, jobs and reports.

  • Loopback or remote HTTPS connections, saved connection preferences and separate desktop sessions per Server address.

  • Uploads, downloads, cookies and WebSocket updates follow the Server protocols.

  • macOS Apple Silicon DMG and Windows x64 installers, with a SHA-256 checksum manifest.

安装与连接 / Install and connect

  1. 先部署或连接 FrameFetch Server。

  2. 在下方 Assets 下载对应系统的 .dmg 或 .exe 安装包。

  3. 默认 Server 地址为 http://127.0.0.1:8111/;远端部署通过 --backend-url=https://your-server.example/ 指定 HTTPS 根地址。

  4. 使用 Server 账户登录;可用业务与 AI 能力取决于服务端配置。

  5. Deploy or connect to a FrameFetch Server.

  6. Download the .dmg or .exe for your system from Assets below.

  7. The default Server address is http://127.0.0.1:8111/. Use --backend-url=https://your-server.example/ for a remote HTTPS deployment.

  8. Sign in with your Server account. Available workflows and AI depend on Server configuration.

源码运行 / Run from source:

git clone --branch v0.2.0-beta.1 https://github.com/StephenQiu30/video-electron.git
cd video-electron
pnpm install --frozen-lockfile
pnpm dev

使用 package.json 指定的 Node.js/pnpm 版本。完整连接与构建说明见 README。

Use the Node.js/pnpm versions specified in package.json. See README for connection and build instructions.

验证与预览范围 / Validation and preview scope

源码固定到 c8a85c94548a20619bf8b34a71a6916992654a17。安装包直接来自该提交的 成功 CI,分别为 FrameFetch-0.2.0-mac-arm64.dmg 与 FrameFetch-0.2.0-win-x64.exe;包内版本及文件名为 0.2.0,GitHub tag 的 beta.1 表示公开预览渠道。

The source is fixed to that commit. Installers come directly from its successful CI build. Package version and filenames are 0.2.0; the GitHub tag's beta.1 identifies the public preview channel.

  • 安装包尚未完成发布者签名,macOS 未公证,系统可能提示安全警告。

  • macOS ARM64 与 Windows x64 自动构建、单元测试、开发态及打包态传输 E2E 通过;这些测试使用受控夹具。干净安装、升级、卸载和真实 Server 完整业务仍需独立验证。

  • 不提供 Intel macOS 或 Linux 安装包。

  • 桌面端需要可达的 Server;媒体处理与 AI 推理由服务端执行。

  • 旧版本本地文件保留原件,不自动导入 Server。

  • Installers are unsigned; the macOS build is not notarized, so system security prompts may appear.

  • Automated builds, unit tests and development/package transport E2E passed on macOS ARM64 and Windows x64 using controlled fixtures. Clean installation, upgrades, uninstallation and complete real-Server workflows require separate validation.

  • No Intel macOS or Linux installer is provided.

  • A reachable Server is required for media processing and AI inference.

  • Files from older local versions remain untouched and are not automatically imported into Server.

FrameFetch 生态 / Ecosystem

欢迎 Star、关注 Releases,或报告系统、客户端版本和 Server 版本相关的使用反馈。

Star the project, watch Releases, or share feedback with your OS, client version and Server version.

首次发布源码 / First-release source · MIT License