[idn] Re: names of various sorts

[Mark.Andrews@nominum.com]

> This message is in MIME format. Since your mail reader does not understand
> this format, some or all of this message may not be legible.
> 
> ------_=_NextPart_001_01BF8CDF.BDCF5390
> Content-Type: text/plain;
> 	charset="iso-8859-1"
> 
> 
> Before you, Mark, started to confuse the issue, it was clear to
> everyone (on this list) that the domain names we were trying to
> internationalise were primarily the ones seen by "ordinary people".
> If other domain names are (or need to be) internationalised along
> with that, then that is a bonus.

	I started to confuse everyone?  There was plenty of confusion
	on this list before I started saying anything.  Even the chair
	did not understand the terms he was using (he does now).

	There were people on this list saying that domain names were
	restricted to 'a' - 'z', '0' - '9', '-' and '.'.  There were
	others saying they were restricted to ASCII.  Yet others that
	know what is actually legal in a domainname and what is the
	standard presentation form.

	Everybody was talking at cross purposes.

	I set out to *remove* confusion by actually reminding people
	that there are terms that describe the subsets of domainnames
	that are talking about.

	The IETF is a technical body.  Everyone here should be capable
	of dealing with technical issues and precise technical
	definitions.

> By the way, your message is an excellent illustration of why
> suggestions like CIDNUC and (the misnamed) UTF-5 (it's NOT a UTF)
> are unacceptable suggestions: even after a decade of QP the
> encoding still leaks through to the poor reader, and it's a
> reencoding into ASCII that is so unique essentially to e-mail
> that no ordinary programs that handle text handles that reencoding.
> The same will be true for CIDNUC and (the misnamed) UTF-5,
> so such proposals should not be pursued further.  We've lived
> through a decade of QP horror and it still haunts us (quite
> needlessly, 8bit IS available) and we certainly don't need
> several more decades of CIDNUC horror.
> 
> 
> 		Kind regards
> 		/kent k
> 
> 
> > -----Original Message-----
> > From: Mark.Andrews@nominum.com [mailto:Mark.Andrews@nominum.com]
> > Sent: Sunday, March 12, 2000 11:36 PM
> > To: idn@ops.ietf.org
> > Subject: Re: [idn] Proposed suggestions from Asia Pacific Top
> > LevelDomain meeting 
> > 
> > 
> > 
> > > 
> > > <smaller>Niklas.=C5berg@=C4lvdalen-skog.se.  And so on for 
> > the worlds
> > > all</smaller>=20
> > 
> > 	
> > 	Because you could not encode the above email address in
> > 	the DNS with then restrictions people on this list seem to
> > 	want to put on international domain names.  The IETF does
> > 	not allow periods in interational domain name labels.
> 
> ?????  Are you really trying to tell us something or are you just
> continuing to try to confuse people?

	No, I was trying to point out what will happen if we don't
	deal with the whole issue.

> If you want to explain
> something and want me to understand, you have to put it in terms
> that somebody who is not very familiar with the details of the
> protocols you are talking about can understand.  Note that
> silly comparisons with apples and oranges DO NOT help,

	Well when terms like 'subset' don't get the message across
	I thought I would try a common metaphor that most people on
	the list would have had contact with in the First Class (K1).

> nor do unexplained acronyms for various kinds of 'records'.

	SRV see RFC2782
> 
> > 	Or sorry you can't a a SRV record to redirect http for
> > 	=C4lvdalen-skog.se to a different set of boxes.  The IETF
> > 	(through the idn working group) was too short sited to
> > 	fully internationalise the DNS, they just internationalised
> > 	host names and underscores are not legal in hostnames.
> 
> I'm not sure what you mean by "http" in connection with "boxes".

	And you can't even look at the example presented and see that
	there is a underscore and http in a label even if you havn't
	looked up what a SRV record is.

> Anyway, it appears to me that HTTP can already handle domain
> names in UTF-8 without any major problems (except that there is
> no normalisation nor case folding).  I.e., apart from that some
> operators are now actively blocking that kind of names. Given
> that that change (to block) only took a few days to implement,
> I guess it can only take a few days to unblock...
> 
> > 		e.g. 
> > 			_http._tcp.\196lvdalen-skog.se
> > 
> > 	Or to put it another way IDNS will not get past the IESG
> > 	and IETF last call unless it deals with all domain names
> > 	and not just host names.
> > 
> > 	Can you now see why I am trying to point out the distinction
> > 	between domain names and host names.  Can you also now see
> > 	why we need to define what is a IDN and what is a IHN.
> >
> > 	Mark
> > --
> > Mark Andrews, Nominum Inc. / Internet Software Consortium
> > 1 Seymour St., Dundas Valley, NSW 2117, Australia
> > PHONE: +61 2 9871 4742                 INTERNET: 
> > Mark.Andrews@nominum.com
> > 
> 
	We are potentially changing the existing DNS. It would be
	polite of everyone on this list if they took the time to
	find out how it is used today so that the proposals being
	made do not break existing functionality that is being used.

	Mark
--
Mark Andrews, Nominum Inc. / Internet Software Consortium
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: +61 2 9871 4742                 INTERNET: Mark.Andrews@nominum.com