RE: Including a 1.1 entity in a 1.0 document

 

> -----Original Message-----
> From: public-xml-core-wg-request@w3.org 
> [mailto:public-xml-core-wg-request@w3.org] On Behalf Of Norman Walsh
> Sent: Monday, 2006 July 17 13:23
> To: public-xml-core-wg@w3.org
> Subject: Re: Including a 1.1 entity in a 1.0 document
> 
> / John Cowan <cowan@ccil.org> was heard to say:
> | Norman Walsh scripsit:
> |> / John Cowan <cowan@ccil.org> was heard to say:
> |> | In that case, the Third Edition of XML 1.0 would 
> control, which specifies
> |> | that the text declaration of the external entity MUST 
> specify version 1.0.
> |> | So it's a well-formedness error.
> |> 
> |> Where do we say that an incorrect version is a WF error?
> |
> | The version number is specified by production 26 as exactly "1.0".
> | Section 2.1 requires that a well-formed document must match
> | production 1 ("document").  So no WFC is needed.
> 
> Ok. I see that. I think it might be worth adding something slightly
> less subtle in some future rev, but I don't think it's urgent.
> 

It is the case that we made production 26 be exactly "1.0"
only in the third edition.  In the 2nd Ed, we had:
  VersionNum 	   ::=    	([a-zA-Z0-9_.:] | '-')+
which wouldn't have made version="1.1" a well-formedness
error.  All we said in the 2nd Ed was:

  The version number "1.0" should be used to indicate
  conformance to this version....
  Processors may signal an error if they receive documents
  labeled with versions they do not support.

But we came out with the 3rd Ed in conjunction with XML 1.1,
so if the universe of discourse includes 1.1, it should 
reference the 3rd Ed (at least) of 1.0 thereby substantiating
JohnC's claim.

We could make that more obvious.  Should we pretend we got a
comment to that effect on the 4th Ed review (that recently
ended) and add such wording to the 4th Ed Recommendation text?

paul

Received on Monday, 17 July 2006 18:34:18 UTC