Karl

> I'm not sure if that is correct. Of the 4 kinds of paths Windows supports (yes, really), one of them is DOS device paths, for example: `\\.\C:\Windows\`. Since it...

> FWIW `file:///C|/a` serializes to `file:///C|/a` on Chrome on Windows, **matching the current spec.** Hm? I still see `file:///C:/a` on the [jsdom live viewer](https://jsdom.github.io/whatwg-url/#url=ZmlsZTovLy9DfC9h&base=YWJvdXQ6Ymxhbms=), and that's still what I would...

FYI @alwinb, I have a lot of additional tests in my implementation of a new URL type for Swift: https://github.com/karwa/swift-url/blob/main/Tests/WebURLTests/additional_constructor_tests.json I'm using a very different path parsing algorithm to the...

FWIW, I've been looking at interoperability between this standard and the `URL` type in Apple's Foundation framework (which I assume would also be of interest to WebKit). It is documented...

I think this is possible, but can be quite tricky. IMO this operation should avoid anything which would cause _other_ URL components to be interpreted with a different meaning, and...

Deprecation without removal is a little difficult, because there's no particular function or public type to annotate. What I could perhaps do is log some kind of warning when we...

> as all the imports are protected with #if canImport(Network) Actually, they are not all protected currently, but this patch adds those conditions. The main `HTTPClient.swift` has a plain import,...

I agree that having dual Package.swift files is unfortunate. I had to do it recently with another project in order to support DocC. You never feel good about it. That...

@weissi I'm a bit busy at the moment working on URLs (news on that coming this week!), and I don't really understand all the ins-and-outs of HTTP so I'm not...

Hmm, well I saw it happen quite a lot. I plumbed the `maxConnections` variable to the `ConnectionPool` initializer so I could tweak it, and from 0-4, everything was fine. At...