[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

RE: Undeclared namespace

• From: Andrew Layman <andrewl@m...>
• To: xml-dev@i...
• Date: Mon, 1 Nov 1999 13:33:24 -0800

Would you seriously like me to go on a campaign to change this?  I can.  But
before you say "yes," though, please make sure that you have a settled,
detailed technical proposal that you are all happy with, is tested, and that
is agreed on by the relevant W3C activities.  If you do that, I will be
happy to ask to have the behavior changed. 

-----Original Message-----
From: Mark Birbeck [mailto:Mark.Birbeck@i...]
Sent: Monday, November 01, 1999 1:19 PM
To: xml-dev@i...
Subject: RE: Undeclared namespace


Andrew Layman wrote:
> Mark Birbeck, regarding the Microsoft MSXML parser handling of xmlns
> attributes in DTDs, wrote "This seems to be generally accepted as
> incorrect."  
> 
> Scanning the archives will reveal that the interaction of 
> namespaces and
> DTDs is one in which every decision has drawbacks. Microsoft 
> took the most
> conservative position.  This may not have been the best course (though
> debate here on XML-Dev and elsewhere does not seem to reach a settled
> technical conclusion), but in any case, by virtue of being 
> conservative, it
> leaves open the possibility of loosening the handling in the 
> future if the
> least-problematic behavior becomes known, whereas the reverse 
> would not have
> been true.  

I was trying to respond to the other messages without re-opening old
wounds Andrew, but since you have ...

I seem to remember the only person who hadn't reached a 'settled
technical conclusion' was you. From my recollection of the debate no-one
was saying how grateful they were that a namespace attribute declaration
had to be #FIXED. The reason for that is obvious; every single DTD that
specifically declares a namespace attribute but doesn't use #FIXED will
cause a validation error. Yet there is nothing in XML 1.0 or the
namespace rec that indicates this should be the case. As is often
pointed out on this list, DTDs know nothing of namespaces. As far as
DTDs are concerned it's just another attribute. If the DTD designer
wants that particular attribute to always have the same value then
that's up to them to declare it, but it shouldn't be built into the DOM!

This whole 'conservative' and 'future loosening' thing is just rubbish.
You make it sound like a minor problem which may or may not be fixed in
the future. But don't you see that the consequence of your developers'
preference has made DTDs over which one has no control completely
unusable? The most significant for me are the three for XHTML.

Mark Birbeck

xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev@i...
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ and on CD-ROM/ISBN
981-02-3594-1
To unsubscribe, mailto:majordomo@i... the following message;
unsubscribe xml-dev
To subscribe to the digests, mailto:majordomo@i... the following
message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa@i...)

xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev@i...
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ and on CD-ROM/ISBN 981-02-3594-1
To unsubscribe, mailto:majordomo@i... the following message;
unsubscribe xml-dev
To subscribe to the digests, mailto:majordomo@i... the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa@i...)

• Follow-Ups:
  • Re: Undeclared namespace
    • From: Paul Prescod <paul@p...>
  • RE: Undeclared namespace
    • From: Mark Nutter <mnutter@f...>

• Prev by Date: Re: A Plea for Schemas
• Next by Date: RE: A Plea for Schemas
• Previous by thread: RE: Undeclared namespace
• Next by thread: RE: Undeclared namespace
• Index(es):
  • Date
  • Thread