So here is what I'm thinking about. Web services making zope a very popular home/intranet application. Using a CMF installation and a few python scripts it is easy to create a place where one can store photos, music, and a weblog. Using the cmf maks everything cataloged so content can be found dynamicly with a search, making linking and organizing things easy and independent of their actual path on the server. Music and photos could be linked by title name.
I think in a envroment like the CMF with a good skin in place there is no need for continued structual site design. Content should be very easy to add. I feel that structured text could be used in such an envroment to tie all all of the binary objects togther. First though it needs a little lift. That is why I've began work on ParsedSTX. It taks its name from ParsedXML becuase the ideas used are similar. StructuredText creates a dom like tree structure when initial parsed. My goal is store that dom in a persistant object so that it is easly accessable. This lets developers access stx documents with dom-like commands. I also wished to add some features to the standard stx HTML renderer. These include automatic generation of id attributes and style attributes for html tags. Also the ability to determin insite links from external links, and do catalog searches on internal links that do not have valid paths. This would allow a person to use the sytax "some inner link":My paper on Plato. and have a catalog search find documents with the title of 'My paper on Plato' and insert the proper path in the href attribute of the link.
The first part of this ramble is becoming true. ParsedSTX is still very broken, but I do have a system in place that lets users blog into the CMF via the blogger api. See RPCAuth and CMFWeblog for more information.