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

Re: RE: Namespaces Best Practice

• From: Jonathan Borden <jborden@m...>
• To: Paul Spencer <paul.spencer@a...>
• Date: Thu, 20 Sep 2001 00:15:11 -0400

Paul Spencer wrote:


> I'm late coming to this thread, but am I the only one to disagree with
point
> 1? (I agree totally with 2.)
>
> Take a typical document submitted to UK Government (my area, by the
> principles apply elsewhere). This will have a document payload that will
> generally meet requirement 1. It will have a single default namespace
> defined in its document element. If it uses other namespaces, the bindings
> will probably be defined here as well. This is then wrapped in an
envelope.

Yes, well this is all fine. First I intend a "best practices" to be used as
a guideline and not as a law of any sort. Second, in XML everything becomes
a "document" so we lose the usage of the term "document" as a letter
contained in an envelope as opposed to the technical usage of the term
"document" to refer to the envelope which contains a letter.

>
> So these "compound documents" definitely should have multiple default
> namespaces defined for different parts of the document. What I would not
do
> is specify bindings at random through the document. I guess that is what
> Jonathan is referring to.
>
yes. again usage of the term "document" has become so overloaded in XML land
that it has become essentially meaningless. The idea of a compound document
is one that itself contains other "sub" documents and so I think it is
entirely reasonable to provide namespace prefix declarations for each of the
"sub" documents if this is desired.

Jonathan

• References:
  • RE: Namespaces Best Practice
    • From: Paul Spencer <paul.spencer@a...>

• Prev by Date: RE: RE: Namespaces Best Practice
• Next by Date: Re: RE: Namespaces Best Practice
• Previous by thread: Re: RE: Namespaces Best Practice
• Next by thread: Re: RE: Namespaces Best Practice
• Index(es):
  • Date
  • Thread