[topicmapmail] XTM Datatypes [Was: Adding weigths to associations]

[Murray Altheim]

Steve Pepper wrote:

> At 18:08 04.12.2002 +0100, Lars Marius Garshol wrote:
> 
>> * Anthony B. Coates
>> |
>> | I would like to be convinced that embedding of an XTM topic map
>> | inside a larger XML document is uniformly supported by TM tools.
>>
>> I think it is uniformly unsupported. After all, XTM does not allow
>> it.
> 
> XTM requires topic maps in the XTM interchange syntax to contain
> a correctly structured <topicMap> element, but it does not require
> that element to be the document element. So topic maps may indeed
> be embedded in other documents - in XTM as well as in HyTM.


The Conformance section of the XTM 1.0 Specification indeed allows
this. It's not hard to support in XML software.


> This is also supported by the Ontopia Knowledge Suite and perhaps
> other topic map engines as well. (Kal? Graham? Jan? Robert?) The
> topic map embedded in the following document loads just fine in the
> Omnigator:
> 
>    <?xml version="1.0" encoding="ISO-8859-1"?>
>    <foo>
>      <bar/>
>      <topicMap xmlns="http://www.topicmaps.org/xtm/1.0/"
>                xmlns:xlink="http://www.w3.org/1999/xlink">
>        <topic id="jill">
>          <baseName>
>            <baseNameString>Jill Hacker</baseNameString>
>          </baseName>
>        </topic>
>        <baz/>
>      </topicMap>
>    </foo>
> 
> Note that "foreign" elements can even be included within the
> topic map element itself (provided validation against the XTM DTD
> is not being performed). They are simply ignored by the XTM reader
> of the OKS.


So you *do* take the Microsoft approach after all. "Simply ignored"

is a lie. There is no "simply ignored" in software. What this really
means is that you've extended an interchange syntax with content
that will operate differently within different applications, so
topic map applications that grok the mysterious, proprietary
content function differently. And if that "differently" is functional
as it likely could be, the user experience of those that use tools
that don't understand the proprietary content suffers. (I don't
know why I'd have to take up this old argument with you, Steve)

This would be a corruption of the concept of interchange; as you
say, it "can be included [...] provided validation against the XTM
DTD is not being performed" (which itself means the document is not
a conformant XTM document). The fourth clause under Conformance
prohibits this:

    *  There are no elements or attributes contained in any of
       the <topicMap> elements that are not explicitly declared
       by the XTM 1.0 DTD specified by this Specification. A
       <topicMap> element is non-conforming if it contains
       elements or attributes from any other “XML Namespaces”
       at any recursive level of element containment.

So if Martin wants to scream, this is I believe this is what he
is yelling about. And apparently what we disagree about changing.

I agree that it's relatively simple to extract <topicMap> elements
from a larger document, and have no problem with this even though
I still would have preferred it not have been allowed. But to
allow other content within the topic map itself would essentially
mean XTM was no longer XTM, but some mix of namespaces, hardly
an interchange any longer but a way for companies to use features
in competition as did Microsoft and Netscape. This did nobody any
good in the end. It didn't create innovation, only balkanization.

Murray

......................................................................
Murray Altheim                  <http://kmi.open.ac.uk/people/murray/>
Knowledge Media Institute
The Open University, Milton Keynes, Bucks, MK7 6AA, UK

            If you're the first person in a new territory,
            you're likely to get shot at.
                                                     -- ma