[topicmapmail] FW: [xfml] XFML as a Topic Map Application?

[Lars Marius Garshol]

* David Steinberg
| 
| How complicating would it be to recast as an application of the
| Topic Map Reference Model (see
| http://www.y12.doe.gov/sgml/sc34/document/0298R1.htm) i.e. make it
| truly a Topic Maps Lite application?  What would be the plusses and
| minuses of doing so?

In general, I think the benefits of this would be that it would make
it clearer what the meaning of XFML information is, in that there
would then be a clean mapping to topic maps. This would almost
certainly help you get buy-in from the topic map community, and it
might also help others trying to learn XFML.

Another benefit is that it would make it much easier to create XFML to
topic map converters, which could be very useful. For example, in
theory the Omnigator might get XFML support.

So I think that in principle this is a good idea, and the topic map
data models were developed precisely in order to allow people to do
this kind of thing.

As Jan says there are other options than just the Reference Model. I
would say that you have the following choices:

  1) Define a mapping to the XTM syntax. This would have the benefit
     that XTM is here now, and that it would be straightforward for
     most people to understand. It might also be a little awkward to
     have to do everything in syntax terms.

  2) Another possibility would be to make a mapping to the Standard
     Application Model[1], similar to those that will be defined for
     the XTM and HyTM syntax. This would probably be easier than
     mapping to XTM (since the SAM is higher-level), and you would be
     able to reuse whatever machinery the XTM/HyTM syntax
     specifications use to make it easy to define such a mapping.

  3) The third way is to map to the Reference Model, as you suggest.
     I would think that this is the most complicated option, but if
     there are XFML features that don't have any clear SAM counterpart
     this may still be the best way to go. 

Also, if you've mapped to the SAM you will automatically be able to
map to the RM, but the converse does not necessarily hold (it depends
on what your resulting RM instances look like).

[1] <URL: http://www.isotopicmaps.org/sam/ >

-- 
Lars Marius Garshol, Ontopian         <URL: http://www.ontopia.net >
ISO SC34/WG3, OASIS GeoLang TC        <URL: http://www.garshol.priv.no >