The Web as an information structure

Pages on the web are files interconnected by links. The pages are stored in a hierarchical structures of folders or directories. This page is in a folder called 'devices' which contains an index file and several folders, each of which contains examples of devices such as a central heating controller, a telephone, and so on.

Each example of a device consists of a main file and perhaps some associated image files.

The index file contains links which point to the main file for each example. That main file, in turn, may contain links which point to the image file(s).

In the HTML code, the links are represented as some magic syntax plus the address of the target. So, the index file will list the addresses of all the main files for the examples. There is no automatic back-pointer.


Apply the CDs to this structure(not to the HTML itself, but the structure of folders, files and links). Compare the profile that for TAG and other notations. Consider viscosity, for example - how much work is required if another entity is to be added? Are there any hidden dependencies? Is premature commitment involved? And so on.