W3C home > Mailing lists > Public > www-tag@w3.org > May 2007

Re: [XMLVersioning] Definition of an XML Language (was: Re: (Partial) review of Versioning XML)

From: John Cowan <cowan@ccil.org>
Date: Mon, 14 May 2007 09:59:17 -0400
To: noah_mendelsohn@us.ibm.com
Cc: www-tag@w3.org
Message-ID: <20070514135917.GC29674@mercury.ccil.org>

noah_mendelsohn@us.ibm.com scripsit:

> I agree with Norm on this.  Furthermore the draft formulation comes close 
> to suggesting that it's the markup that is definitive of the XML language. 

I think the difference is terminological.  XHTML is commonly referred
to as a language, even though there are various abstract schemas (1.0
Transitional, 1.0 Strict, 1.1, etc.) that specify particular sets of
constraints on it.  It seems useful to me to standardize a term such as
"language" (better than "tag set") for the element and attribute names
and attached semantics independent of how they may be arranged.

(I say "abstract schemas" because of course 1.0 Transitional can be
specified in many ways in many different schema languages, from the
original DTD to W3C XML Schema, RELAX NG, or TagSoup Schema Language.)

> As Norm says and part 1 strongly suggests , an XML 
> language is a class of texts, all of which are constrained by the rules of 
> XML 1.x to be at least well formed, and some of which are further 
> constrained for the purpose of conveying some particular sort of 
> information (technical report, purchase order, TAG Finding, whatever.)  I 
> think that part 2 should discuss the versioning and evolution of such 
> languages.

If the definition of "XML language" is that stringent, then there
is no real concept of change: there are just more and less similar
XML languages.  1.0 Transitional and 1.0 Strict are distinct languages
that happen to resemble each other somewhat.  But in what aspects are
they similar?  Element and attribute names and associated semantics,
precisely.  Introducing a separate notion of "XML language" allows a
level of abstraction which matches the way people actually use the term.

-- 
John Cowan  cowan@ccil.org  http://ccil.org/~cowan
Any sufficiently-complicated C or Fortran program contains an ad-hoc,
informally-specified bug-ridden slow implementation of half of Common Lisp.
        --Greenspun's Tenth Rule of Programming (rules 1-9 are unknown)
Received on Monday, 14 May 2007 13:59:31 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 26 April 2012 12:47:45 GMT