Re: ISSUE-59: normative-language-reference FPWD

On Jan 23, 2009, at 13:34, Sam Ruby wrote:

> In yesterday's call, a number of issues were raised when the topic  
> of making "HTML 5: The Markup Language" available First Public  
> Working Draft.

I am opposed to publishing "HTML 5: The Markup Language" as a REC  
track draft (or on an unspecific track but claiming normativity). This  
view also represents a broad enough consensus at Mozilla.

We'd be OK with publishing it as a draft on track to become a Note  
that gives an informative view to a subset of the subject matter over  
which "HTML 5" is normative.

The specific objections are the following:

1) "HTML 5" already normatively defines the HTML5 markup language. It  
doesn't make sense for the working group to compete with itself by  
publishing two normative documents about the same to thing. (Resolving  
this by removing the overlap from "HTML 5" would not be an acceptable  
way to resolve this, since that would lead us back to bad old  
situation where the "language" and the processing model are in  
different specs with a risk of things falling through the cracks.)

2) "HTML 5: The Markup Language" makes a schema normative. Experience  
with HTML4 shows that normative schemas freeze innovation and  
competition in validator development, because one implementation is  
declared normative and improvements on the implementation are  
considered wrong. For years, HTML4 validation wasn't improved at all., Validome and Relaxed have now improved on things but are  
still seen as illegitimate compared to DTD-based HTML4 validation. If  
the document is frozen at a certain snapshot in time (as a REC would  
be frozen), validators either couldn't improve or would have to  
deviate from the normative schema at the risk of being perceived  
illegitimate. A schema is code--a part of a validator implementation.  
WGs don't make particular snapshots of C++ code normative, either.

Henri Sivonen

Received on Friday, 23 January 2009 13:32:59 UTC