Hi, I have the initial elements of a database design for a Content
Management System, and would like some feedback before committing to it
and progressing further. The diagram can be accessed at
http://www.intraspin.com/db.gif.

Articles can be of various types, including XHTML 'free' content,
documents, etc. They can also be of type 'rule' to select other
articles based on the tags an article is associated with. Articles
make up the content of the site.

The hierarchical navigation of the site will be based on Tags. The
tree is defined in the 'Tag Relations' Table. Each Tag may or may not
form part of this hierarchy, based on whether it is part of the tree.
Articles are assigned a primary tag, to allow them to be 'placed' on a
'page'. They will also be assigned a plethora of other tags (including
but not limited to those between the primary tag and the root of the
tree). Thus, articles of type rule will be able to pull other articles
in to pages that aren't those articles' primary page.

There is thus a many-to-many relationship between Articles and Tags,
such that a tag can be applied to many articles, and and article can
have many tags. There is also a many-to-many relationship between Tags
and Tag Parents, allowing a Parent to have many children, and allowing
a child to have more than one parent.

I would appreciate any feedback as to what problems you foresee and
their gravity (i.e. to what extent can the problem be solved).
Problems might include efficiency.

Many thanks,
Iain