[pmwiki-users] hierarchical groups, revisited

[Bellave Jayaram]

> -----Original Message-----
> Categories offer the beginning of this story, but I think 
> there is more to be offered.  I'm not sure exactly what's 
> there to be implemented, but I think page attributes and XML 
> are probably involved.
> 
> Jon
> 
> [1] - http://www-dimat.unipv.it/biblio/isko/doc/colon.htm
> 

I think the bigger context here when talking about hierarchical groups is
the information architecture that is possible to be realized on a PmWiki web
site.

When working with PmWiki, I tend to think as if the Group maps to the
Personality facet (in colon classification terminology (CCT)). Thus,
anything dealing with Animals (whether it be Matter, Energy, Space or Time)
has its own page within that group - Categories could be used for each of
the facets to further group the pages. I would consider Canines also to be a
Personality facet, have a group called Canines and so on. If one is familiar
with the information architecture of the site, then one can construct
hyperlinks in various ways, add groups/pages etc. The problem arises when
such information is not available to a user or the site evolves in a ad-hoc
way and it is not possible to impose strict architectural guidelines.

If one were using Farms and Fields, then one can certainly consider the Farm
as the Personality and each Field as one of M, E, S, T and Groups for
others. Categories could be used for any facet or isolate (CCT) but I know
there are some limitations in PmWiki with inter-farm, inter-field and
inter-group support when it comes to Categories - I don't have enough
knowledge about this. All this is only addressing subject classification
(DITA provides some support for this - see [1]). However, there also needs
to be some facility for handling topic types (or infotypes in DITA
terminology) in PmWiki. I thought there might be a way to do this using
template pages and there probably is but I am still lost......

IMHO, Jon makes a very good suggestion. If something can be implemented and
XML is considered, I hope it will be based on a standard such as DITA or
OWL.

Jayaram

[1] http://www.ibm.com/developerworks/xml/library/x-dita10/