W3C home > Mailing lists > Public > www-tag@w3.org > December 2012

Re: The polyglot specification should be suitable for normative reference

From: Henry S. Thompson <ht@inf.ed.ac.uk>
Date: Thu, 06 Dec 2012 12:57:58 +0000
To: Noah Mendelsohn <nrm@arcanedomain.com>
Cc: "www-tag\@w3.org" <www-tag@w3.org>, Henri Sivonen <hsivonen@iki.fi>, lachlan.hunt@lachy.id.au, Eliot.Graff@microsoft.com, Norm Walsh <ndw@nwalsh.com>, Maciej Stachowiak <mjs@apple.com>, Paul Cotton <Paul.Cotton@microsoft.com>, Sam Ruby <rubys@intertwingly.net>
Message-ID: <f5b624f2v8p.fsf@calexico.inf.ed.ac.uk>
Noah Mendelsohn <nrm@arcanedomain.com> writes:

> I am convinced by the argument that Tim Berners-Lee made on our TAG
> teleconference last week that "normative" describes not the status of
> an individual document, but of a reference from one document to
> another. 

I have to disagree.  Certainly a normative reference functions to
expand the normative content of a standard by incorporating the
normative content of the referred-to spec. into the referring spec.
But this observation doesn't address the question of what it means to
_be_ normative content.

I think there are two main cases:

 1) Standards which describe artefacts: protocols, languages, physical
    objects, etc.  In this case normative content determines
    conformance: conformant messages, implementations, instances,
    etc. are those which satisfy all applicable normative
    requirements.  Such requirements are often, although not always,
    expressed using RFC2119 [1] vocabulary.

 2) Standards which define vocabulary or other standards
    infrastructure (for example, notation).  In this case it is the
    normatively defined vocab/notation/... which is available for use
    in referring specs.  The relevant definitions accordingly need not
    use RFC2119 language, although they sometimes do: A normative
    definition of the form "To be a 'framis', an artefact *must* have
    four legs" is in practice no different from one of the form "A
    'framis' is defined to be an artefact with four legs", given that
    conformance will come from some referring spec. including,
    normatively, something such as "The result *must* be a 'framis'
    [ref. ...]"

So for example the Infoset spec., which defines terminology, doesn't
itself define conformance, and the XPath Data Model spec [3] makes
normative reference _to_ the Infoset spec., in normative conformance
statements, for example:

  "The Infoset must not contain any *unexpanded entity reference
   information items* [ref. XML Infoset]."

The net result is a requirement on processors which claim conformance
to the XDM.

Although I'll address some of the questions around the 'polyglot'
spec. in a subsequent message, from the perspective advanced here the
simple fact that a standard might well make normative reference to it,
in order, for example, to normatively require polyglot input, is
itself sufficient for it to make sense for the 'polyglot' spec. to
contain normative content.

ht

[1] http://www.ietf.org/rfc/rfc2119.txt
[2] http://www.w3.org/TR/xml-infoset/#conformance
[3] http://www.w3.org/TR/xpath-datamodel/#const-infoset
-- 
       Henry S. Thompson, School of Informatics, University of Edinburgh
      10 Crichton Street, Edinburgh EH8 9AB, SCOTLAND -- (44) 131 650-4440
                Fax: (44) 131 650-4587, e-mail: ht@inf.ed.ac.uk
                       URL: http://www.ltg.ed.ac.uk/~ht/
 [mail from me _always_ has a .sig like this -- mail without it is forged spam]
Received on Thursday, 6 December 2012 12:58:53 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 6 December 2012 12:58:53 GMT