Request for early feedback: Fetch server API
还没有人认领这个 Issue。
- 主要语言
- JavaScript
- 星标
- 122k
- 派生
- 37.3k
- 平均合并
- 4 天 2 小时
- 30 天内合并 PR
- 283
描述
As part of the new unified http API strategic initiative, I am working on a new WinterTC/TC55 proposal for a new "Fetch Server API".
A key goal of this is a server API that is runtime agnostic; and in particular not-Node.js specific. It also covers all versions of http (h1, h2, h3).
You can find the rendered draft here: https://proposal-fetch-server.jasnell.me/
Github repo for issues/discussion here: https://github.com/jasnell/proposal-fetch-server
Some background reading for ideas leading into this:
- https://www.jasnell.me/posts/http-server-api-draft
- https://www.jasnell.me/posts/fetch-needs-error-codes
- https://www.jasnell.me/posts/fetch-is-not-enough
- https://www.jasnell.me/posts/quic-comes-to-node
Note: it is well established that the current standard definitions of the fetch Request/Response/etc APIs are not entirely suitable for server side use given a strict reading of the specification. There's no reason to rehash those arguments. A goal of this work is to address those limitations.
This is NOT YET an official TC55 work item. It is a draft proposal that I will be presenting to the committee at the next meeting. Before then, I'm soliciting feedback on the design.
@nodejs/quic @nodejs/http @nodejs/http2 @nodejs/net
贡献指南
从这里开始
- 先读完整个 Issue,再读项目的贡献指南。
- 在 Issue 下留言说明你要接手 —— 这能避免两个人做同样的事。
- Fork 仓库,在一个分支上完成修改。
- 提交 Pull Request,并在描述里引用这个 Issue 编号。
调研方向
先阅读渲染后的 Fetch Server API 草案和链接的 proposal repository,然后查看背景文章以了解设计背景。在提议于 TC55 进行展示之前,为与 runtime 无关的 API 设计提供可执行的反馈,即视为完成。
由索引模型根据 Issue 内容生成。
评估
- 领域
- api, networking
- Issue 类型
- 功能
- 难度
- 5/5
- 预计耗时
- 一周以上
- 活跃度
- 冷清
- 描述清晰度
- 需要澄清
- 新手友好度
- 20/100