Is there any way we can visibly elevate our finally-chosen term to some kind of revered status like the "MUST" and "SHOULD"s of the world? What I mean is something that will cause any reader to subconsciously recognize every time they encounter the term (when used in the normative way) that this term Has A Significant Meaning whose definition they should be sure to seek out and understand before reading any further.
Maybe mark it up with <term> or some such thing that would end up visibly styled in the published spec? (At least in some forms if not all.)
This might help to avoid reader confusion in the long run, no matter which choice we end up with.
FWIW, my feelings strongly echo Bob's from start to finish, except that I was going to say "schema definition language".
On top of the baggage already mentioned, "schema" all by itself has the very unfortunate additional baggage of being confused with "W3C Schema".
From: firstname.lastname@example.org [mailto:email@example.com] On Behalf Of Bob Thomas
Sent: Thursday, March 13, 2014 7:21 AM
To: Eliot Kimber
Cc: Kristen James Eberlein; DITA TC
Subject: Re: [dita] Looking for best generic language to refer to DTD, XSD, and RELAX NG in the DITA 1.3 spec
I believe that using "schema" would be the least surprising for most of our readers. "XML schema language" would be more accurate, but it would be cumbersome to work into the wording giving it a legalistic tone that occludes meaning.
Even though I prefer "schema" for the reader's sake, my first choice is for the specification is "grammar". Eliot's concerns, about schema connoting a degree of specificity that isn't present, are what convince me "grammar" is a better choice. There is simply too much computer-science baggage attached to the word "schema".
Whatever we decide needs to be documented in the specification's Terminology section.
[Eliot, sorry for the redundant message. I meant to reply to all in the first one you received.]
On Thu, Mar 13, 2014 at 7:27 AM, Eliot Kimber <firstname.lastname@example.org> wrote:
I think "grammar" is probably the most technically correct generic term,
in that that's what DTDs, XSDs, and RELAX NG grammars actually define:
declarative syntactic constraints on the construction of XML documents,
which is analogous to the grammar of a language, that is, the rules for
constructing valid instances, regardless of actual meaning.
I think the term "schema" carries too much database weight and implies a
level of typing that is not present in DTD or RELAX NG and only weakly
defined in XSD.
Nevertheless, "schema" is probably the term that most readers would
I used the phrase "document-type constraint language" because it's more
inclusive than "grammar" or "schema". In particular, procedural tools like
Schematron can do more than simple grammar or lexical constraint: they can
define population constraints, meaning checks on the specific details of
the document content, such as ensuring that if you specify text X in
element Y that you must also have element Z with text W, but only on
alternate tuesdays or months with no full moon. So that goes beyond the
concept of a simple declarative schema or grammar.
Eliot Kimber, Owner
On 3/13/14, 8:05 AM, "Kristen James Eberlein"
> With the advent of support for RELAX NG, we either need to insert
> RELAX NG everywhere in the spec that we refer to "DTD and XSD," or
> we need to find a good, generic alternative.
> Here is a typical paragraph that I would like to revise using a
> generic term; I've highlighted in r
> the places that I want to make generic (or have attempted to make
> " DITA does not require that conforming DITA documents have an
> associated document-type definition as long as all required
> attributes are explicit in the document instances. However, most
> DITA documents have an associated DTD,
> RELAX NG, or XML Schema document. Such associated
> documents enable validation using normal XML processors; they also
> can provide default values for the @domains and @class attributes.
> While the DITA specification only defines coding requirements for
> DTD, RELAX NG, and XML Schema documents, conforming DITA documents
> MAY use other document-type constraint
> languages, such as Schematron."
> What is the best choice for an generic term? "XML schema language"
> or simply "schema"?
> I also want to replace the adjective "constraint" in the last
> sentence; I've highlighted it in blue bold.
> Any suggestions?
> Kristen James Eberlein
> Chair, OASIS DITA Technical Committee
> Principal consultant, Eberlein Consulting
> www.eberleinconsulting.com <http://www.eberleinconsulting.com>
Time zone: Mountain (GMT-7)