RE: <check> Response Attribute

["Hollenbeck, Scott" <shollenbeck@verisign.com>]

> -----Original Message-----
> From: Eric Brunner-Williams in Portland Maine
> [mailto:brunner@nic-naa.net]
> Sent: Monday, September 24, 2001 2:47 PM
> To: Hollenbeck, Scott
> Cc: ietf-provreg@cafax.se
> Subject: Re: <check> Response Attribute 
> 
> 
> So now the text is controlling, not the schema?

I'm not sure how you inferred the above from my last message, which cited
the current schema.  Just to be clear, I'll repeat what I said last Friday
[1] and what I meant in my last message [2]:

The schema is the formal specification.

Section 4.2 of epp-04 (the shared structure schema) enumerates two possible
values for the "x" attribute: "+" and "-".  The text in section 2.9.2.1
reads as follows:

... "and contain an "x" attribute whose value identifies the object as
either "+" for a known object or "-" for an unknown object."

I do not see any inconsistency between the schema and the text in terms of
describing the two valid values.  If your contention is that the meaning of
the two values is described in the text, and not the schema, you're quite
correct -- that's a limitation of the current schema, and that's exactly why
I think both schema and text need to be changed.

This thread started with a suggestion to replace what I initially wrote as a
kludge to provide a binary yes/no response in as few characters as possible
with a standard XML Schema method of expressing binary alternatives.  The
original suggestion has been met with both disagreement and agreement with
an amendment (amendment: add an additional attribute to describe reasons for
unavailability).  I didn't think this would be such a contentious issue, but
given the debate I'd like to ask for chair guidance before either keeping
things as-is or making any changes.

<Scott/>

[1] http://www.cafax.se/ietf-provreg/maillist/2001-09/msg00099.html

[2] http://www.cafax.se/ietf-provreg/maillist/2001-09/msg00120.html