IDNA comments

John C Klensin klensin at
Tue Jul 8 00:08:37 CEST 2008

--On Monday, 07 July, 2008 14:55 -0700 Mark Davis
<mark.davis at> wrote:

> For historical reasons, English does have a special place in
> DNS. The purpose of IDNA is to erase that distinction as much
> as possible, allowing characters needed by any language. The
> point here is not about that general issue, but that *not
> every character that people thinks are part of words in a
> language* is necessarily going to be in IDNA. And as an
> example of that, even English in its 'special' position does
> not get perfectly ordinary, common words like "don't".


I want to think further about whether I can come up with good
constructive examples, but I tend to agree with Frank that the
identification of contractions as words is a little dubious and
hence that these are not the most wonderful examples.  I do note
that the name of the 50th state (the one in the middle of the
Pacific, for those for whom that is a better reference) cannot
be spelled correctly in ASCII, much less as an LDH string.
There are also several words in English that were imported from
other languages, mostly French, for which the preferred
spellings in most authoritative dictionaries retain diacritical
marks or other decorations.   Better examples might lie in those

Still studying your other comments.


More information about the Idna-update mailing list