On Jan 25, 2005, at 1:36 PM, Ray Denenberg, Library of Congress wrote:
> In fact with separate namespaces it's even easier than that -- in a
> instance, or in a mads instance, you use just <name>, no prefix -- the
> default namespace. ( In an instance of a third schema referencing one
> the other than it would use the appropriate prefix.) So at least for
> example separate namespaces seems to be a simplifying rather than
> complicating factor. (Yes it's a bit more complicated because we're
> about factoring out the common part to the type library and that part
> be namespaced.)
I didn't quite get this, so am not sure I'm on the right track below.
> Well, from the example above, flexibility to define names without fear
> collision. For whom: those defining names and those creating complex
> instances that reference them.
I'm looking for concrete examples. Why should I want to create
instances with mixed MODS and MADS namespaces?
This is critical issue. For anyone who's ever worked with instance
data across multiple namespaces, every time you add a namespace -- in
the context of XSLT processing, or query, or whatever -- the level of
complexity goes up significantly.
I spent ages and ages utterly frustrated trying to work out multiple
namespace issues in some XSLT stylesheets I was working on awhile ago.
In that case, I had the following namespaces:
bib metadata (mods)
styling config language (my own schema)
[I've since added a fourth which is internal to the processing; and
then of course there's the output namespace]
So, in this fairly complex context, I had three namespaces, and it was
a PITA. It was made somewhat easier, however, because each document
had its own namespace.
Now, if I understand you right, you're telling me it's feasible I may
need to deal with one or two more namespaces just to be able to handle
MODS data? If I have that right, why I should I want to make my life
more difficult (having to constantly be aware of which element is
associated with which namespace) to get what practical benefit?
If one doesn't get the namespaces precisely right in XML processing,
it's as if the data isn't even there.