Imagine a book that told a different story every time it was opened. The story might change depending on the gender of the reader, or the sex. It might depend on the location of the reader, or the position of the book in time; the time of day, or time in years. Centuries might pass before the book tells the same story again.
The nature of the web makes such a book possible. Immediately, a simple reading of the user-agent to determine the reader’s operating system and browser could be used to present each with a different version, breaking the narrative along several general pathways. Sections could be hidden or revealed by simple manipulation of the layout.
Secondly, parsing the IP address of the reader would reveal their rough geographical location, or the institution they were calling from. In the first instance, sentences could be run through rough online translators, translating passages into – or out of – the reader’s assumed language. Different nations could be offered different political perspectives on the narrative. In the second, those from academic institutions would find appended a wealth of sources, some true, some false, while government agents might find the entire pages reduced to Xs and punctuation marks.
Finally, simple randomisation could alter the meaning of certain words, their tense or number. Names would be changed, emphasis misplaced. But random number generators are no such thing, and each has a pattern. A one time pad.
The final stage attempts to preclude the existence of a master copy.