Re: [radext] I-D Action: draft-ietf-radext-nai-02.txt

[Alan DeKok <[e-mail]>]

Stefan Winter wrote:
> Then I'd suggest making this very explicit. E.g. replace the previous
> sentence with
> 
> "See [RFC5198] and section 2.6 of this specification for a discussion of
> normalization. Strings which are not in Normal Form Composed (NFC) are
> not valid NAIs and SHOULD NOT be treated as such."

  Looks good to me.

> I don't think servers are the main problem when it comes to
> normalisation (or just valid UTF-8 in the first place). I have much more
> concerns about ignorant supplicants.

  I agree.  It would be nice to support feedback in one of the EAP
methods (e.g. TEAP).  A server could say "Thanks for using DOMAIN', but
next time please use DOMAIN".

> With the multitude of end-user computing devices in the wild, and the
> varying amounts of care firmware writers spend into their supplicant
> code, I'm sure there'll be a whole zoo of representations of
> internationalised domain names on the supplicant side :-(

  Yes.  Which is why I'd like to get the NAI document published.  RFC
4282 is distressingly broken.  Being able to point to a *good* document
would be useful.

> Would it be okay to *RECOMMEND* that UTF-8 which is not in NFC should be
> normalised prior to "doing something" in terms of lookup in the config?

  Yes.

  Alan DeKok.