Re: [idn] Dots, and a path to working IDNs

["D. J. Bernstein" <djb@cr.yp.to>]

Eric A. Hall writes:
> So that we know exactly what it is

``The long-term IDN solution will encode Unicode characters as UTF-8 on
the wire.'' I suspect that we have consensus on this point, even though
we don't have consensus on the short-term approach.

> that we'd be voting for or against

I'm sure several people will pop up to inform you that IETF has ``straw
polls'' but doesn't have ``votes.'' Votes are things done by legitimate
standards organizations that obey United States antitrust law; the IETF
is not one of those organizations.

> How do you prevent utf8 chars from breaking legacy devices?

Fix the legacy devices. This is obviously something that's going to
happen anyway.

> an older (and orphaned) X terminal that legitimately rejects
> underscore in hostnames

Are you fantasizing, or can you identify an actual X terminal that has
to be fixed?

Same question about your ``hundreds if not thousands of other
applications that ... reject data that does not comply with LDH rules.''
Do you have any evidence that the fixes listed on my web page---in
particular, enabling no-check-names in /etc/resolv.conf, with the
current gethostbyname() library---are not sufficient?

> I don't think you can do it with the current protocol

You are speaking nonsense. The proposal changes the protocols.

> and implementation limitations.

Implementation problems are a serious issue. That's why I've been going
to so much effort to identify the implementation work that actually has
to be done. See http://cr.yp.to/proto/idn.html.

---Dan