[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [idn] I-D ACTION:draft-ietf-idn-idna-08.txt
Mark.Andrews@isc.org wrote:
> The alphabetic characters in this context are 0x41-0x5a and 0x61-0x7a
> (zero parity ASCII A-Z, a-z). All others are compared exactly.
That may be what the popular DNS implementations do, but I don't see RFC
1035 saying how to compare 0x80-0xFF.
"Eric A. Hall" <ehall@ehsco.com> wrote:
> You are the only one that is unclear on it. 0x80-0xFF must match
> exactly.
I'm not the only one. Here's an excerpt from a message by John Klensin:
http://www.ietf.org/mail-archive/ietf/Current/msg15742.html
RFC1035> For all parts of the DNS that are part of the official
RFC1035> protocol, all comparisons between character strings (e.g.,
RFC1035> labels, domain names, etc.) are done in a case-insensitive
RFC1035> manner. At present, this rule is in force throughout the
RFC1035> domain system without exception.
John Klensin> To emphasize, that is "all parts" and "all comparisons",
John Klensin> not "unless you happen to find the high bit turned on".
John Klensin> An existing and conforming implementation has no way to do
John Klensin> those required case-insensitive comparisons outside the
John Klensin> ASCII range.
Robert Elz> No, nor is it required to.
John Klensin> There we probably disagree -- I suggest that the text is
John Klensin> at least ambiguous and might require it.
AMC