fix(core): make webpack HMR always listen to current location

[jeengbe]

See #7912 for explanation as to how these changes fix the problem. This does not have any side effects that I am aware of.

Pre-flight checklist

• I have read the Contributing Guidelines on pull requests.
• If this is a code change: I have written unit tests and/or added dogfooding pages to fully verify the new behavior.
• If this is a new API or substantial change: the PR has an accompanying issue (closes #0000) and the maintainers have approved on my working plan.

Motivation

Test Plan

Test links

Deploy preview: https://deploy-preview-_____--docusaurus-2.netlify.app/

Related issues/PRs

Closes #7912

[facebook-github-bot]

Hi @jeengbe!

Thank you for your pull request and welcome to our community.

Action Required

In order to merge any pull request (code, docs, etc.), we require contributors to sign our Contributor License Agreement, and we don't seem to have one on file for you.

Process

In order for us to review and merge your suggested changes, please sign at https://code.facebook.com/cla. If you are contributing on behalf of someone else (eg your employer), the individual CLA may not be sufficient and your employer may need to sign the corporate CLA.

Once the CLA is signed, our tooling will perform checks and validations. Afterwards, the pull request will be tagged with CLA signed. The tagging process may take up to 1 hour after signing. Please give it that time before contacting us about it.

If you have received this in error or have any questions, please contact us at cla@fb.com. Thanks!

[netlify]

✅ [V2]

Built without sensitive environment variables

Name	Link
🔨 Latest commit	9c3497f
🔍 Latest deploy log	https://app.netlify.com/sites/docusaurus-2/deploys/62fddb423536ac0009a8ad34
😎 Deploy Preview	https://deploy-preview-7951--docusaurus-2.netlify.app
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.

---

To edit notification comments on pull requests, go to your Netlify site settings.

[github-actions]

⚡️ Lighthouse report for the deploy preview of this PR

URL	Performance	Accessibility	Best Practices	SEO	PWA	Report
/	🟢 94	🟢 98	🟢 100	🟢 100	🟠 80	Report
/docs/installation	🟢 90	🟢 100	🟢 100	🟢 100	🟢 90	Report

[facebook-github-bot]

Thank you for signing our Contributor License Agreement. We can now accept your code for this (and any) Meta Open Source project. Thanks!

[slorber]

This does not have any side effects that I am aware of.

Can you backup your claim with some reference to other projects using a similar setup?
If nobody uses this by default, there must be a reason 🤷‍♂️

I'd rather align to the behavior others are already using if we were to merge something not easy to double-check is correct for everyone.

For example CRA has env variables, but does not use such default:
https://github.com/facebook/create-react-app/blob/main/packages/react-scripts/config/webpackDevServer.config.js
facebook/create-react-app#7750

[jeengbe]

This does not have any side effects that I am aware of.

Can you backup your claim with some reference to other projects using a similar setup? If nobody uses this by default, there must be a reason 🤷‍♂️

I cannot, I just can't come up with a situation in which the new behaviour would be unwanted. As the websocket runs on the same domain + port the rest of the website is on, it always ought to be accessible on the same domain + port.
The only environment in which this would cause an issue is a reverse proxy that does not work with websockets, and the original port still exposed, and I do not see that happening in any dev envionment I can think of.

[Josh-Cena]

Not familiar with the dev server architecture at all, so more of a question: does this change the experience with remote developing, like on GitHub codespaces? Right now we have to recommend --host 0.0.0.0.

[jeengbe]

It does change the experience in that it fixes HMR for when Docusaurus is launched on a different address/port is is being used with. --host 0.0.0.0 is still necessary for the server to work at all, but this pr eliminates issues with the socket connection (for HMR) being made to a different address/port (since setting to 0 makes it use the current location's)

[Josh-Cena]

The current configuration looks sensible to me (sounds like a no-op in most cases). I'll be willing to give it a try.

[slorber]

Let's try this and we'll see if anyone complains. So far it seems to work for me.

For now I won't backport it in 2.x

[zevisert]

I for one would like to see this backported to 2.x! I brought up a question in discord that was likely solved by this PR. https://discord.com/channels/398180168688074762/867060369087922187/1012096405780844594

My use case is the same as #7912 -- it's that we want to run the docusaurus dev server on eg minikube locally. One of our requirements is to have TLS enabled in development for secure-context only JS features, like navigator.clipboard - and we plan on doing that with an ingress that terminates the TLS connection. We can get our SSL certs by proving a domain like local.example.com that points at 127.0.0.1.

Problem is that when deployed behind a reverse proxy ingress, the HMR websocket tries (and fails) to connect to the internal port like wss://local.example.com:3000/ws instead of wss://local.example.com/ws. I confirmed that @docusaurus/core@canary (0.0.0-5272 as of time of writing) solves this.

[slorber]

@zevisert we could backport it but I'd prefer to have more user feedback first. For example, if nobody complains on any 3.0 upcoming alphas, it becomes safer to backport it

[slorber]

@zevisert we could backport it but I'd prefer to have more user feedback first. For example, if nobody complains on any 3.0 upcoming alphas, it becomes safer to backport it

[jeengbe]

Until the backport fix is out, remember that you can always put PR diffs in your patches folder and run patch-package.

https://patch-diff.githubusercontent.com/raw/facebook/docusaurus/pull/7951.diff

[masad-frost]

I noticed this never made it to 2.2.0, why?

[slorber]

because we don't have a lot of feedback on potential side-effects due to v3.0 canaries not being widely, and I decided to not backport this:

#7951 (comment)

If you want to use it, wait for 3.0, use a canary release, or this temp workaround: #7912 (comment)

[mi-na-bot]

I am trying to use a 2.4 project in GitHub Codespaces, which sets up an HTTP proxy for port forwards, and this issue is a major disappointment/pain. One zero-side-effect option would be to allow the user to enable this with the CLI. Having the dev server listen on the same host/port as the original connection and the client component look for it there is exactly what Next.js has been doing for quite some time with no ill effect. IMO, the current configuration only works because of a side effect.

[mi-na-bot]

Applying the change from this PR in my node_modules directory fixes the issue in GitHub Codespaces.