[Upstream #121] 卡在了Exception in handleNewProject: Network Error

[ivanzud]

Upstream Issue Mirror

• Upstream issue: 卡在了Exception in handleNewProject: Network Error 666ghj/MiroFish#121
• Upstream repository: 666ghj/MiroFish
• State: closed
• Author: jackytianjp
• Labels: question
• Last updated: 2026-03-11T02:15:10Z

Summary

Local Coverage

• Status: covered
• Summary: New-project startup errors now resolve raw frontend Network Error failures into an actionable backend URL and proxy/CORS hint instead of leaving users at a generic handleNewProject failure banner.
• Local refs: frontend/src/api/errors.js, frontend/tests/errors.test.mjs, https://github.com/ivanzud/MiroFish/issues/41
• Notes: Mirrored into fork issue [Upstream #121] 卡在了Exception in handleNewProject: Network Error #41 on March 11, 2026. This backfills the landed frontend error-diagnostics work already represented by upstream PR fix: improve new-project network error diagnostics 666ghj/MiroFish#125.

Recent Upstream Comments

• dosubot[bot] at 2026-03-10T15:49:04Z: Hi @jackytianjp! I'm Dosu and I’m helping the MiroFish team. 这个 "Network Error" 发生在 [handleNewProject](https://github.com/666ghj/MiroFish/blob/985f89f49acbb44ee14d9d680682c…
• darfy22 at 2026-03-10T17:17:48Z: 我也是遇到同样的问题，会不会是因为部署在云端服务器的原因，后端服务配置地址需要在哪里改为服务器IP地址？ @dosu
• jackytianjp at 2026-03-11T02:15:02Z: ## ✅ 已解决 感谢Dosu的帮助和启发 ### 问题描述 使用 Docker 部署时，前端调用 /api/graph/ontology/generate 接口报 Network Error，后端日志中完全没有收到任何请求。 ### 根本原因 前端代码默认将后端地址写死为 http://localhost:5001。在 Docker 容器内部这没有问题，但当你通过 NAS 或服务器的 IP 地址（如 http://192.168.x.x:3000）访问前端…