encoding VS url

Be aware that with the WHATWG Encoding specification [1], that says that latin1, ISO-8859-1, etc. are aliases of the windows-1252 encoding, not the proper latin1 encoding. As a result, browsers and operating systems will display those files differently! It also aliases the ASCII encoding to windows-1252.

[1] https://encoding.spec.whatwg.org/#names-and-labels

They used to support UTF16 (both LE and BE) before it got removed because no one uses UTF16: https://github.com/whatwg/encoding/issues/18

Semicolon (;) has no special meaning in a URL. You can ascribe it a meaning in your particular routing, but the spec has nothing to say about it.

https://url.spec.whatwg.org/

https://github.com/whatwg/url/issues/577#issuecomment-118534...

It's not insurmountable absolutely and I would appreciate it absolutely.

For the query portion, it really depends if your are reading it server side or client side and using the WHATWG standard[0] which itself just mirrors convention. However, the standard dictating how a URL might be formed does not mandate anything about the query string that makes it parsable.

0]: https://url.spec.whatwg.org/#interface-urlsearchparams

Browsers have discrepancies too of course. Here's an interesting Chromium bug I've been following: https://bugs.chromium.org/p/chromium/issues/detail?id=125253... and an associated WHATWG discussion: https://github.com/whatwg/url/issues/606

Some multiple examples of browsers disagreeing: https://www.yagiz.co/url-parsing-and-browser-differences

What's imho missing is a size or count method. The reason they don't have one yet is because it's not clear whether it should be all tuples, or all keys. Discussion

Well, [text](href) is just a lousy syntax. Quite apart from how easy it is to forget which way round it is, the way round that it is is syntactically inferior: the parentheses are URL code points <https://url.spec.whatwg.org/#url-code-points>, so URL encoding won’t percent-encode parentheses, so Markdown doesn’t actually support all valid URLs, leading to injection attacks if all you do is regular URL encoding, deliberate or accidental, and deciding where an href ends is troublesome and inconsistent, with some Markdown implementations terminating at any right parenthesis, and others trying to match parentheses as a heuristic that helps most cases. The other way round, with the href in square brackets, would have been better in this regard, as square brackets aren’t URL code points, and thus will be percent-encoded. But better still would have been to lean on angle brackets more, matching long-held custom and the other style of links Markdown already uses (just plain ). In my own lightweight markup language that I’ve been working on for a while and am now polishing up and implementing properly, I’m currently using [text ]. [text] is also quite tempting, with slightly different trade-offs.

(When I speak of the details of URL encoding and which characters get percent-encoded, these things weren’t quite so clearly-defined back in 2004 as they are now, but I believe it was all still true.)

>A domain is a non-empty ASCII string that identifies a realm within a network. [RFC1034]

>The example.com and example.com. domains are not equivalent and typically treated as distinct.

https://url.spec.whatwg.org/#concept-domain