graphql-over-http VS fetch

Another major pain is the fact that operation names and HTTP codes are often hidden in request bodies which makes it really hard to see what is going on in standard monitoring tools. It's possible to write some converter but it's beyond me why some people decided that requests which failed on the server respond with status 200. GraphQL spec doesn't define how it should behave, there's only a proposal (https://github.com/graphql/graphql-over-http)

The GraphQL over HTTP specification states the following:

However, if you are doing GraphQL subscriptions over Server Sent Events (HTTP) (which is currently not part of the GraphQL over HTTP specification), the data flow is only from server to client. So each operation must be a separate request, which should be no problem when using HTTP/2, as the browser connection limit is not hit so fast (There are also workarounds to this if you cannot ise HTTP/2).

Complementary to the note, you can learn more about GraphQL over HTTP in the specification over here: https://github.com/graphql/graphql-over-http/blob/main/spec/GraphQLOverHTTP.md

GraphQL-spec, GraphQL-over-HTTP: guarantees your GraphQL API to work with all existing GraphQL clients (Apollo, Relay, URQL, and more).

Note: While GraphQL can be done over almost any protocol, this article focuses on the most commonly used protocol GraphQL over HTTP. However, most knowledge can be transferred to other protocols such as GraphQL over WebSockets or other more exotic ones.

There is also the GraphQL over HTTP Specification repository https://github.com/graphql/graphql-over-http

graphql-sse is a reference implementation of the GraphQL over Server-Sent Events Protocol aiming to become a part of the GraphQL over HTTP standard.

With no further ado - I humbly introduce graphql-ws. A coherent, feature-full, zero-dependency, plug-n-play, lazy, simple, server and client implementation of the new, security first GraphQL over WebSocket Protocol with full support for all 3 GraphQL operations: Queries, Mutations and Subscriptions. The protocol aims to be standardised and become a part of GraphQL with the help of the foundation’s GraphQL over HTTP work group.

You could additionally also choose a specific HTTP code for that scenario (although that would contradict the official GraphQL over HTTP spec https://github.com/graphql/graphql-over-http/blob/main/spec/GraphQLOverHTTP.md).

mTLS has warts when used cross-origin. Fetch spec says that pre-flight requests mustn't include client certificates[1], so as a consequence servers behind mTLS authenticated proxy won't get a chance to reply to those pre-flight. Yet for non-preflighted requests it's fine to include client certificates..

[1] https://fetch.spec.whatwg.org/#cors-protocol-and-credentials

I've been testing full duplex streaming from and to the browser using fetch() in a Native Messaging host. (No browser currently support full duplex streaming even though HTTP/2 does, see Fetch body streams are not full duplex #1254).

Most js libraries use XMLHttpRequest and so provide HTTP_X_REQUESTED_WITH: XMLHttpRequest, but neither Chrome's implementation nor Github's polyfill of the new fetch uses a similar header. So how can one detect that the request is AJAX?

Any resource of significance should be given a URI. https://www.w3.org/DesignIssues/Axioms.html#uri

Or alternatively,

> Cool URLs don't change (implicitly, cool things have URLs, see above). https://www.w3.org/Provider/Style/URI

The advantage would be so high. It'd become a standard way to assert a resource, to make known a fact, that would be viable across systems. Instead of pushing to a chat app an anonymous chat message in a room, the server could assert a /room/42/msg/c0f3 resource, could identify universally what it is it's sending.

We have come glancingly close to getting such a thing so many times. The HyBi mailing list that begat websockets had a number of alternate more resourceful ideas floating around such as a BEEP protocol that allowed patterns beyond request/response of resources. The browser actually implements an internal protocol that uses HTTP2/push to send resourceful messages... Even though http2/push was de-implemented for webserving in general, and even though ability to hear push events was never implemented (oft requested).

The best we have today is to stream json-ls events, which have an @id property identifying them. But developers would have to snoop these events, and store them in a service worker, to make them actually accessible as http resources.

I continue to hold hope eventually we'll get better at using urls to send data, to assert new things happening... But it's been nearly 30 years of me hoping, and with some fleeting exceptions the browser teams have seemed disinterested in making urls cool, in spite of a number of requests. https://github.com/whatwg/fetch/issues/65 was an old request. https://github.com/whatwg/fetch/issues/607 had some steam in making it happen.

Fetch

Proposal: fetch with multiple AbortSignals - I got the idea of merging multiple signals from here.

Port 10080 is blocked on most browsers[0] per the WhatWG "bad ports" list[1]. That particular port was added to the list due to the Slipstream attack[2] that made the news a few years ago[3].

You don't have to switch to a browser that ignores standard security mitigations. Just pick a different port for your service.

[0] I just tested Chrome, Firefox, and Safari.

[1] https://fetch.spec.whatwg.org/#bad-port

[2] https://samy.pl/slipstream/

[3] https://news.ycombinator.com/item?id=24955891

Note that caching resources across sites isn't really a thing anymore. See https://github.com/whatwg/fetch/issues/904