[Bug 9178] add definition of "conforming document"

http://www.w3.org/Bugs/Public/show_bug.cgi?id=9178

Noah Mendelsohn <nrm@arcanedomain.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |TrackerRequest

--- Comment #11 from Noah Mendelsohn <nrm@arcanedomain.com> 2010-10-27 15:31:05 UTC ---
Thank you for the attention to my concerns.  Unfortunately, I don't feel that
the response in http://www.w3.org/Bugs/Public/show_bug.cgi?id=9178#c10
adequately addresses my concerns.  

For a language like HTML5, I think it is important and valuable to have a good
story about whether a given document is "HTML5-compliant".  The approach taken
here says, basically, "it's in the eye of the beholder".  I do understand that
line of thinking, I just think it's not for the best.

First of all, we're in agreement that it should be OK for "applicable
specifications" to be written that extend or change the HTML5 language.  Where
we seem to disagree is on 1) whether it's worth having a clean, rigorous
definition of what it means to be compatible with the unextended specification
and 2) [optionally] whether there's standard terminology suggested for
expressing that a document or piece of software is compatible with "HTML5 as
extended with specification XXX".

I think that at least (1) is very important.

> Noah Mendelsohn wrote:
> 
> > Wouldn't it be better to require that such documents be
> > referred to as "conforming to HTML5 as modified by my-malicious-spec-X" (or in
> > the more likely example "conforming to HTML5 as modified by
> > my-nonmalicious-spec-that-makes-significant-and-perhaps-otherwise-incompatible-changes"? 

Ian Hickson wrote:

> What would such a requirement mean?

Again, my main point is that it would mean that you cannot use the term
"HTML5-compatible" to refer to compatibility with the extended specification,
only for documents or software that conforms to unextended HTML5.

> How would you test it? What conformance class
> could pass it? What happens if someone decides
> that that requirement doesn't apply to them?

I believe the HTML5 draft already attempts to define the main conformance I
care about, which is to the unextended specification.  

I believe your question is "how would you test..." conformance to something
like "HTML5 as extended with Noah's specification"?  I think the answer is to
say something along the lines of:

"Where applicable specifications are used to extend or change HTML5, those
specifications SHOULD define the conformance criteria for the extended
language."

Example 1: Compatible extension
===============================

Let's say I write a simple specification that augments HTML 5 by requiring one
additional attribute to be used by some document tracking system I'm promoting.
 I write Noah's Doc Management spec, which defines the augmented language, to
require the attribute.  I would in the extension specification write:

"For a document to conform to this specification, it must be a conforming HTML5
document, with the additional requirement that the @noahdoctrack attribute MUST
be specified on the <HTML> element."

Example 2: Incompatible extension
=================================

Now let's say I introduce an incompatible change (for some reason). Noahs
Specification Number 2 for some bizarre reason requires mismatched quotes on
attributes (I'm only doing this to show how incompatible changes would be
handled.)

The specification would say:

"For a document to conform to this specification, it must be a conforming to
all the conformance requirements for HTML5 documents, except with regarding the
quotation of attributes.  Attribute values MUST be quoted with a leading single
quote and closed with a double quote.  Thus, documents conforming to
HTML5+Noah's specification #2 will also be conforming HTML5 documents if and
only if the documents contain no such attributes."

This shows the style in which incompatible changes could be introduced.

------------

Overall, I think this terminology is practical, simple, and useful. More
importantly, I think it gives much stronger meaning to saying that a document
is "HTML5 compatible", and that's very important.

Ian Hickson wrote:

> If you would like to escalate the issue to
> the full HTML Working Group, please add
> the TrackerRequest keyword to this bug,
> and suggest title and text for the tracker
> issue

I am adding the TrackerRequest keyword.

The title I suggest is: Conformance terminology

Also: I am at this point raising the issue myself, not on behalf of the TAG as
a whole.  Again, thank you for your consideration of my concerns.  

Noah

-- 
Configure bugmail: http://www.w3.org/Bugs/Public/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the QA contact for the bug.

Received on Wednesday, 27 October 2010 15:31:19 UTC