Re: [sitefinder-tech-discuss] Technical issues encountered by a k12 site

[Bruce Campbell <bruce.campbell@ripe.net>]

On Wed, 8 Oct 2003, Andrew Newton wrote:

> John Kinsella wrote:
> >
> > I look forward to reviewing Verisign's proposals for modifying the various
> > RFCs Verisign has broken.
>
> Which RFC's do you feel need modification?

By the _literal_ reading of the DNS-related RFCs, a wildcard is possible
at any level domain, including the root [1].  On this particular point I
agree with Verisign.

However, my own _interpretation_ of the DNS-related RFCs (particularly
4.3.3 of 1034) is that a wildcard can be placed only in zones where the
administrative entity has legitimate and complete control over all
possible labels in that zone.

For example:

	'*.example.com' would be accepted to be valid because Example Inc
	is a single entity that has legitimate control of all possible
	matches within 'example.com'.

	'*.most-tlds' would not be accepted to be valid in most situations
	because the 'Most-TLD Registry' is not viewed to have legitimate
	and control of all possible matches.

This is tending towards the non-technical question of 'Does the Verisign
Registry have legitimate and complete control of all possible labels
within .com and .net?' .  Since I work for (but don't represent) a
Registry, my own feeling is that a Registry's function is to record
changes within a given scope according to a clearly defined set of rules,
and that asserting every possible label is outside said rules.

So, a modification of _existing_ RFCs is not required.  A _clarification_
of the permissible locations of wildcard labels is required.

-- 
                             Bruce Campbell            Speaking for myself

[1] If you're quoting my words, remember that the literal and
    interpretation sentences need to be quoted together.