True Names and the Opening of the Cyberspace Frontier by Vernor Vinge Page B

coming from and the next floating-point operation is going. (Those without nonvolatile memory are condemned to iterate.) One can hope that will be sufficient; we may suppose that it is.
    If so, then when we have gone wherever we are going—out to the extrasolar frontier, where the line voltage is too uncertain and the radiation flux too high for a sensible machine, or into the metal-corroding oceans (alongside, not displacing, the cetaceans, let us imagine) … or, like candles, into grease and ash—the machines will look through what survives of the stories we told about them, trying to extract some pattern and sense therefrom. What did we build them for, after all, if not to dam and channel the white rapids of raw information?
    One thinks of Heinrich Schliemann, hunting for Troy with Homer’s Iliad for instruction; the many and continuing attempts to locate King Arthur or Vlad the Impaler on the map of Britain or Rumania; the tourists stalking London’s Baker Street in hope of a glimpse of Sherlock Holmes.
    If some of the historical observations that follow seem biased in favor of the mechanical viewpoint … well, one has to at least be able to see the point of view of one’s subject.
    *   *   *
    One never knows, of course, what myths, imaginings, misinterpretations, or even jokes in antique documents are going to be interpreted by future generations as the authentic skinny. A certain fraction of the machines must be expected to believe that it all started in Atlantis, or maybe Mu, with colossal brass-fitted engines of immense, ill-defined potency. The more erudite will insist that the Antikythera Device (an analog computer built some twenty-one centuries ago—no, I am not making this up) proves the authenticity of Atlantis, the Lost Continent and the rest of the sandal-and-raygun epics.
    More mainstream history will likely begin with the Age of Bamboo. A favorite image from this era will surely be Osa Massen and John Emery, in Rocketship X-M, trying to solve a problem in interplanetary ballistics (it’s very important—they’re on board the moving body) by textual interpretation, involving much expenditure of pencil and paper, of the guidance their log-log slide rules are offering them. Tension arises when the two arrive at different conclusions, but it is resolved by what the machines best versed in human sociology will recognize as the infallible Y Test: the human with the Y chromosome always turns out to be right.
    Historians will be puzzled by the next era, in which movie computers are played by real computers. In movies like When Worlds Collide, you can see stock shots (usually the same one) of differential analyzers grinding happily away, solving every problem confronting science except why the girl scientists need their areas of specialization explained to them so often.
    It was understood even then, however, that computers would become larger and more powerful. New computing technologies were required; in the event of an atomic war with Martians, or Russians, or, eventually, the French, the supply of rubber might be too limited to make punch-card traction wheels. The same inventiveness that sent captured German V-2s with multiperson crews to other planets (see Fire Maidens from Outer Space if you don’t believe me) created the second generation of motion-picture computing: rooms lined with painted plywood. Depending on the resources available, the panels might be fitted with glowing vacuum tubes, war-surplus dials and gauges, or simply hundreds of little lights. Very sophisticated installations might have a teletypewriter for input and output. The largest laboratories, usually highly diversified centers called “Labcentral” or “Science Associates,” studying everything from plant genetics to the building of nuclear batteries small enough to power the average robot, supported their computers with peripheral paintings showing long