Hi, Well, the document could say less about the topic, but then why is any of it in there now?
No one seemed to claim there was anything incorrect in this PR: https://github.com/richsalz/draft-ietf-uta-rfc6125bis/pull/88 But the feedback seems to be that the WG doesn't want to say this for some (unstated?) reason. It seems very unproductive to briefly reference IDNA2008 only, because that is not how the internet works, even if we wish it were the case. > > how does this knowledge help implementers to properly implement matching IDNs against the names in certificates, which is performed using A-labels? That's in the PR, right? You have the text below, and a brief explanation of why this problem exists. I think the PR is more informative than -10 or saying nothing. "A P-label differs from an A-label in that both types of labels contain valid output of the Punycode algorithm, but P-labels may fail the other quality checks prescribed by {{IDNA-DEFS}} for A-labels." This seems like a good thing to know, but I guess we'll have to wait for the WHATWG to document it all in one place. That's a bummer to me. š thanks, Rob On Thu, Feb 16, 2023 at 4:48 AM Valery Smyslov <smyslov.i...@gmail.com> wrote: > Hi Rob, > > > > Hi, no dispute on the consensus call. > > > > the consensus call is over. Based on the discussion on the mailing list > > the chairs believe that the consensus is to keep the current (-10) text > > and not to go into the details of explaining the current far-from-ideal > > state of arts in the area of Internalized Domain Names. > > This is not true. The -10 draft does go into the details. It just doesn't > do it very well. The references all stay the same in either version, with > the exception of adding the ICANN resolution. > > > > The current text has four parentheticals in one paragraph. The proposed > text is plain spoken and links to the right ICANN document. Basically, > "people break the rules to use emojis, and also use old IDNA standards." > > > > I don't have an opinion on the CABF stuff, but it seems like that is what > actually works. So, this document is inaccurate because it doesn't take a > descriptive approach as it stands. > > > > The problem (as the chairs see it) is that if the document > accurately describes all the complexities of the current state of arts in > this area, > > then it will require quite a lot of new text with very little > (if any) benefit for implementers. Itās cute to know that people do break > the rules and use emojis, > > but how does this knowledge help implementers to properly > implement matching IDNs against the names in certificates, which is > performed using A-labels? > > The goal of this document is not to improve (or even just to > describe) the situation with IDNs, so the chairs believe that the WG > decided to stay apart > > from these complexities and to say as little as possible about > them. > > > > I agree, that the -10 is probably not ideal and if the authors > manage to improve it (along the lines āsilence is goldā), > > for example using Viktorās proposals from his last e-mail, it > will be great. > > > > Regards, > > Valery. > > > > > That said, we think that some points raised during discussion > > may need to be addressed before the document is sent to the IESG [...] > > > > Yes, this is all correct. > > > > thanks, > > Rob >
_______________________________________________ Uta mailing list Uta@ietf.org https://www.ietf.org/mailman/listinfo/uta
