Re: No Namespace Variables and Functions Extensions

Yes, this issue is similar, but its not quite the same.  This issue only 
addresses functions, but not variables.  Also, if your definition of a 
"user defined" variable or function being one which was defined by a 
user (as opposed to the XQuery implementation), then it ignores an 
implementor's ability to polute the default namespace in a negative manner.

Will the resolution you refer to in the next WD address these issues?

Paul Cotton wrote:

>This area of functionality is partially covered by Issue 228:
>
>228. default-namespace-functions: Should we keep the default function namespace, and the xf: namespace?
>http://www.w3.org/TR/xquery/#default-namespace-functions 
>
>This issue was still marked as open in the May 2 drafts since the XML Query and XSL WGs had different interim resolutions to this issue.  The WGs have now resolved their differences and this issue will be resolved in the next public working drafts.
>
>/paulc
>Chair, XML Query WG 
>
>Paul Cotton, Microsoft Canada
>17 Eleanor Drive, Nepean, Ontario K2E 6A3
>Tel: (613) 225-5445 Fax: (425) 936-7329
>mailto:pcotton@microsoft.com
>
> 
>
>
>  
>
A vendor that implements XQuery is free to extend the standard F&O by 
adding new functions and variables as it likes.  It would seem natural 
to me that a vendor who does this would want to create a special 
reserved namespace containing these extensions.  However, my 
understanding is that the vendor could also add these into the no 
namespace as well.

Is this a good idea?  The potential problems I see are two fold:

    * Interoperability suffers because one vendor may have reserved
      variables or functions another doesn't
    * Compliance with the W3C recommendation is questionable since the
      normative examples & use cases provided may or may not work
      "as-is" without some manipulation (changing variable or function
      names)

The analogy I would make is to Java.  Java has a specification and there 
are several vendors who implement that specification.  It would be like 
a vendor would be able to introduce special reserved variable names - 
without a package name.  Yes, there may be value-added, but does this 
JDK conform to the Java standard?

I'm still trying to get my hands around what the W3C wants to mandate, 
what it wants to recommend, and what it wants to leave up to the 
implementation.  One alternative would be to say that the no namespace 
is "final" in that it is closed for extension by XQuery parsers, but 
open to XQuery developers.

I can see it both ways, so I wanted to hear what others thought . . .