2 ms·
Ahh, that is because that isn't a valid URI - it is an IRI. The space of IRI implementations is quite sparse, mostly limited to browsers (and often rather spot
by dwaite 4y ago
Ahh, that is because that isn't a valid URI - it is an IRI.
The space of IRI implementations is quite sparse, mostly limited to browsers (and often rather spotty within the browsers as well). So a page hyperlink or paste into a browser bar will turn it into a URI behind the scenes, but it is unlikely another program or script will handle them.
In fact, I believe WHATWG defines that hyperlinks are URI, not IRI - so the generated HTML in this case by HN is not valid.
IRI are of dubious value - they exist in a space where users are expected to understand and evaluate the legitimacy of an identifier by sight. That said, this is about the best we have for web domains, so (for example) I think all major browsers do render internationalized domain names in their address bar.
- Dylan16807 4y agoScrew that nomenclature. URI is universal. Banning non-ASCII letters doesn't make sense. I will continue to call it a URL/URI.
- zerocrates 4y agoBeyond the usual ASCII stuff, the standard allows you to have all nonsurrogate characters between U+00A0 and U+10FFFD in the path of a URL within HTML. The standard has browsers internally percent-encode them. It's also dumped the IRI/URI terminology and just uses URL.