re: [ciq] An important question between Schema and DTD

[Ram Kumar <rkumar@msi.com.au>]

David,

> Message text written by Ram Kumar
> >
> I do not know how we are going to support this feature in DTD. Is there
> any good document that talks abouthandling namespaces in DTD? Please
> advise.
> This is a very important issue if we want to move forward with namespace
> support for our standards.
> 
> <<<<<<<<<<<<<<<<<<<
> 
> Ram,
> 
> The OAG has been looking at this too.  There users have reacted strongly 
> against prefixes to tags as being really tough on interoperability.
> 
> Instead most people recognize it is easier and makes more sense to
> compartmentalize content within a parent tag and allow the XPath
> context to define the use.
> 
> I.e.
> 
> <Billing>
>      <Address/>
> </Billing>
> <Delivery>
>     <Address/>
> </Delivery>
> 
> Also - this avoids the issue of namespace parsing generally.
> 
> I'm hoping V2.0 of XML will solve this by canning namespace, except
> as an annotation hint, and taking the better <grammer> approach that
> RELAX uses.

Do you have any views on how much of the work of RELAX is expected to be
adopted by V2.0 of XML? Do you know when they expect to release V2.0 and
what is coming in V2.0?

> Also notice that namespace really is an impediment to re-use of content,
> so if I receive :
> 
> <Billing xmlns:foo="somelocation" >
>    <foo:Address/>
> </Billing>
> 
> and then I want to write this out to:
> 
> <Custinvoice>
>     <Address/>
> </Custinvoice>
> 
> then I have a bunch of needless parsing to strip off the foo: part of all
> the tagnames.
> 
> That's just a simple example.  Where multiple namespaces occur within
> a block of content it becomes a nightmare to figure out which is valid
> where in context when inheriting definitions from different schema parents.

Agreed.

> To me this all has AVOID written all over it.   I'm strongly of the opinion
> that namespace is a vaccination for a disease that few people have 
> been found to actually contracted - and where the vaccine is far worse
> than the disease itself.

Whatever I read about namespaces, it says that the W3C were in a rush to 
introduce namespaces without clearly having a business view/impact in mind 
and completely forgot about DTD and its usage of namespaces. The real issue 
is how to handle two or more XML documents that share common names. For 
example, I am working on a client project that takes multiple XML documents 
from many sources outside the organisation and is trying to use them. Name 
conflicts is
a big issue here. They want to keep it simple, but it appears to be not that 
easy. 

Regards

Ram
> DW.