Lists Home |
Date Index |
At 04:07 PM 6/2/2002 -0400, Glenn Marcy wrote:
>In reading this thread main thing that I see confusing people
>is the name of the parameter "qName". I remember when it was
>first something like "rawName", which wasn't as well aligned
>with terms in the Namespace spec. However, it was perhaps a
>slightly less confusing name with Namespaces turned off.
I'd call the change to qName an unfortunate historical accident at this
point, especially given than common advice when using namespace-aware
processing is simply to ignore the qName entirely.
>"dual nature" of the methods in the interface are unfortunate.
>With the Namespaces feature turned off, the only information
>to pass to the application is the "literal name" specified.
>With Namespaces turned on, there are the namespace name and
>the local part parameters to report, and the option that the
>"literal name" (which is a QName in the Namespaces spec) is
>also of interest for some applications that want to access
>the prefix. The "qName" parameter is this literal name from
>the document, which is a slightly odd name for reporting that
>information when Namespaces is turned off. However, that is
>the parameter that is intended to be used for passing this
>information. I find it quite confusing that people want to
>duplicate this information in the interfaces if Namespaces
>is turned off and that they want to do so using yet another
>parameter that has a logical meaning only when Namespaces is
>turned on !!
I'm not sure why you find it confusing or difficult. I find the identifier
"QName" completely and utterly meaningless in a namespace-unaware
environment. QName doesn't appear at all in XML 1.0. "Local name" doesn't
appear either, but I don't believe that it's as alien a concept.
"Every day in every way I'm getting better and better." - Emile Coue