Re: [whatwg/fetch] Browsers preserve fragments on redirects (#505)

@wanderview @jakearchibald @asuth @youennf I'd like to finally solve fragments in URLs and service workers. As far as I can tell we have these scenarios (not sure if synthetic really matters):

```
Document Service worker responds with Outcome
Fetch x  Synthetic     x
Fetch x  Fetch y      y
Fetch x  Fetch y#y     y???
Fetch x#x Synthetic     x#x
Fetch x#x Fetch y      y#x
Fetch x#x Fetch y#y     y???
```
Are there scenarios I missed here?

What I'm observing in Firefox (I find it hard to test in Chrome and Safari due to certificate overrides being a bit of a pain) is that request's fragment is everything. So the first `y???` would be `x` and the second would be `y#x`. This is different from what #696 has (which is the PR from 2018 (that I think got abandoned due to me forgetting about it after vacation) rebased) in that the response is not authoritative when it comes to the fragment. Which I think is a bit weird with how we generally treat service worker responses and the fact that with redirects the location URL fragment is authoritative if present. I no longer care strongly though so I'll defer to you on what to specify here.

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/whatwg/fetch/issues/505#issuecomment-777409833