Re: [idn] W.G. Last Call on "Requirements of Internationalized Domain Names"

["Allen Smith" <easmith@beatrice.rutgers.edu>]

On Feb 13,  9:47am, Marc Blanchet wrote:
> This is a idn working group last call for comments on advancing the 
> following document as a Informational RFC:
> 
>   	Title		: Requirements of Internationalized Domain Names
>   	Author(s)	: Editors Zita Wenzel, James Seng
>   	Filename	: 
> http://www.ietf.org/internet-drafts/draft-ietf-idn-requirements-04.txt
> 
> Please send substantive comments to the idn mailing list, and minor
> editorial comments to the authors.  This last call period will end two
> week from today on February 28th, 2001.

I see a problem with the following section:

>[2.6] The protocol MUST work for all features of DNS, IPv4, and
>IPv6.  The protocol MUST NOT allow an IDN to be returned to a requestor
>that requests the IP-to-(old)-domain-name mapping service.

It's _way_ too limited. It isn't just applications requesting
hostnames from IP addresses that'll get broken if they get back a
name containing characters that they don't expect (namely anything not 
allowed in RFC1123). It's an application asking for _any_ type of DNS
record containing a hostname. I suggest considering the problems on
UNIX, for instance, of a script that calls another program with a
domain name that it got back from an MX request. What happens if that
domain name has shell-active characters (redirection, etcetera)?
Potential problems, up to and including security holes.

Suggested rephrase:

[2.6] The protocol MUST work for all features of DNS, IPv4, and
IPv6.  The protocol MUST NOT allow an IDN violating [RFC1034] and
[RFC1123] to be returned to a requestor that does not request
IDNs. This provision MUST be enforced by either nameservers or
resolvers and SHOULD be enforced by both.

	-Allen

-- 
Allen Smith				easmith@beatrice.rutgers.edu