New section about IDNs #128
Conversation
✅ Deploy Preview for bp-i18n-specdev ready!
To edit notification comments on pull requests, go to your Netlify site configuration. |
|
This seems like the wrong advice. For specifications dealing directly with domains we'd want them to reuse https://url.spec.whatwg.org/#host-parsing. At some point that ends up using IDNA2008 by way of UTS46, but with a lot of caveats. |
|
Sorry, I just saw this. We can change the text to something like this:
What is the reason for not using IDNA2008 directly? Is the rationale recorded somewhere? |
|
IDNA2008 is not what's actually implemented. UTS46 has quite a bit to say about it, which is referenced from URL. |
Co-authored-by: Addison Phillips <[email protected]>
| <p>IDNA2008 and IDNA2003 do not match what's implemented in web browsers. Browsers use IDNA2008 by way of [[UTS46]].</p> | ||
|
|
||
| <div class="req" id="punycode"> | ||
| <p class="advisement">Specifications SHOULD NOT refer to Punycode, especially as a type of string. The IDNA2008 terminology of "A-labels" and "U-labels", defined in RFC 5890, SHOULD be used.</p> |
There was a problem hiding this comment.
Hmm no? They should use the URL terminology and algorithms. Punycode is completely abstracted away.
There was a problem hiding this comment.
Now the problem is that there are some specs that do use the term "Punycode" to refer to the ASCII-Compatible Encoding form. Any suggestions for which term in URL should be used? Such as:
There was a problem hiding this comment.
HTML is being fixed in whatwg/html#10522.
CSP should reference URL's domain to ASCII.
RDF seems like it wants "domain to Unicode", but it also references IETF's URI/IRI so I'm not sure it can be helped.
Fix #119.
This is a new section about IDNs, without much content. Any comments would be appreciated.
Preview | Diff