One of the reasons I wanted to avoid all this and keep it simple. Dino > On Jul 9, 2024, at 10:15 AM, Tim Bray <tbray@xxxxxxxxxxxxxx> wrote: > > On Jul 9, 2024 at 10:07:44 AM, "Salz, Rich" <rsalz=40akamai.com@xxxxxxxxxxxxxx> wrote: >> UTF8 is upward compatible with ASCII. That is, all seven-bit ASCII characters are valid UTF8 characters. > > I don’t want to be pedantic here (sigh, anything involving Unicode always becomes pedantic) but Rich, it wouldn’t actually be a good idea to require “UTF-8”. What you want to require is UTF-8 encoded Unicode characters, and probably not all of them. PRECIS in RFC8264 provides a fully-thought-through subset of Unicode for identifiers. If that’s too complicated, https://www.ietf.org/archive/id/draft-bray-unichars-09.html, now under AD sponsorship by Orie, provides a much simpler but still useful subset. -- last-call mailing list -- last-call@xxxxxxxx To unsubscribe send an email to last-call-leave@xxxxxxxx
