Clock of the Long Now

The project was conceived by Danny Hillis[?] in 1986 and the first prototype of the clock began working on December 31, 1999, just in time to display the transition to the year 2000. At midnight on New Year's Eve, the date indicator changed from 01999 to 02000, and the chime struck twice, to ring in the (popular) second millennium. That prototype, approximately two meters tall, is currently on display at the Science Museum in London.

Design

The basic design requirements of the clock were:

• Longevity: The clock should be accurate even after 10,000 years, and must not contain valuable parts (such as jewels, expensive metals, or special alloys) which would encourage looting.

• Maintainability: Future generations should be able to keep the clock working with nothing more advanced than Bronze Age tools and materials, if necessary.

• Transparency: The clock should be understandable without stopping or disassembling it; no functionality should be opaque.

• Evolvability: It should be possible to improve the clock over time.

• Scalability: To ensure that the final, large, clock will work properly, smaller prototypes must be built and tested.

Obviously, no clock can have a guaranteed lifetime of 10,000 years, but some clocks are designed with guaranteed limits. For example, a clock that shows a four-digit year date will not display the correct year after the year 9999. With continued care and maintenance the clock could reasonably be expected to display the correct time for 10,000 years.

Whether a clock would actually receive continued care and maintenance for such a long time is debatable. Hillis chose the 10,000-year goal to be just within the limits of plausibility. There are technological artifacts, such as fragments of pots and baskets, from 10,000 years in the past, so there is some precedent for human artifacts surviving this long, although no human artifact has been continuously tended for more than a few centuries at most.

Many options were considered for the power source of the clock, but most were rejected due to their inability to meet the requirements. For example, atomic energy and solar power systems would violate the principles of transparency and longevity. In the end Hillis decided to require regular human winding. This may seem an odd choice, but remember that the clock design already assumes regular human maintenance.

The options considered as sources of timing for the clock included a pendulum (inaccurate over the long term, and requires lots of ticks), balance wheel (even more inaccurate), a torsion pendulum (potentially slower, but even less accurate), water flow (inaccurate and wet), solid material flow (inaccurate), daily temperature cycle (unreliable), seasonal temperature cycle (imprecise), tidal forces (difficult to measure), earth's rotating inertial frame, (difficult to measure accurately), stellar alignment (unreliable because of weather), solar alignment (unreliable because of weather), atomic oscillator (not transparent, difficult to maintain), piezoelectric oscillator (not transparent, difficult to maintain), atomic decay (difficult to measure precisely), wear and corrosion (very inaccurate), rolling balls (very inaccurate), diffusion (inaccurate), tectonic motion (difficult to predict and measure), orbital dynamics (difficult to scale), tuning fork (inaccurate), pressure chamber cycle (inaccurate), inertial governor (inaccurate), and even human ritual, (too dependent on humans).

A suitable timing source must be reliable, meaning that it wouldn't easily become stopped. But it must also be accurate. Hillis concluded that that no single source of timing would meet the requirements, so the clock will use an unreliable but accurate timer to adjust an inaccurate but reliable timer, creating a phase-locked loop. Specifically, the current design uses solar alignment to adjust a slow mechanical oscillator, based on a torsional pendulum. The combination can, in principle, provide both reliability and long-term accuracy.

Displaying the time and date

Many of the usual units displayed on clocks, such as hours and calendar dates, are likely to have little meaning after 10,000 years. On the other hand, every human culture counts days, months (in some form), and years. There are also longer natural cycles, such the 26,000-year precession of the earth's axis. On the other hand, the clock is a product of our time, and it seems appropriate to pay some homage to our current arbitrary systems of time measurement. In the end, it seemed best to display both the natural cycles and the some of the current cultural cycles.

The center of the clock will show a star field, indicating both the sidereal day, and the 26,000-year the precession of the zodiac. Around this will be a display showing the position of the sun and the moon in the sky, as well and the phase and angle of the moon. Outside this will be the ephemeral dial, showing the year according to our current Gregorian calendar system. This will be a five-digit display, indicating the current year as something like 02000.

Options considered for the part of the clock that converts time source to display units include electronics, hydraulics, fluidics, and mechanics. A problem with using a convention gear train is that gears necessarily require a ratio relationship between the timing source and the display. The required accuracy of the ratio required increases with the amount of time. For instance, for short period of time in may suffice the count 29.5 days per lunar month, but over 10,000 years the number 29.5305882 is a much better choice. Achieving such precise ratios with gears is possible, but awkward. Instead the millennium clock uses binary digital logic, implemented mechanically. In effect, the conversion logic is a simple digital computer (more specifically, a digital differential analyzer[?]), implemented with mechanical wheels and levers, instead of the more usual electronics. The computer uses a 27-bit number representation, with each bit represented by a mechanical lever or pin that can be in one of two positions. There are about 300 bits in the machine.

Another advantage of the digital computer over the gear train is that it is more evolvable. For instance, the ratio of day to years depends on the earth's rotation, which is slowing at a noticeable, but not very predictable rate. This could be enough to throw the phase of the moon, for example, off by a few days over 10,000 years. The digital scheme allows the number to be adjusted if the length of the day changes in a different way than expected.

Support

The project is supported by the Long Now Foundation, which also supports a number of other very long-term projects, including The Rosetta Project[?] (to preserve the worlds languages) and the Long Bet Project[?]. The Long Now Foundation has purchased a mountaintop near Ely, Nevada, surrounded by the Great Basin National Park, for the construction of the full sized clock.

Alexander Rose was Hillis's primary collaborator on the first prototype clock. The other members of the design team for the first prototype were David Munro, Elizabeth Woods and Chris Rand. Musician Brian Eno is collaborating with Hillis on the writing the music for the chimes for a future prototype.

External links

• Pictures of the prototype and various parts can be seen at http://www.longnow.org/

Bibliography

• Stewart Brand, "The Clock of the Long Now: Time and Responsibility". Basic Books, 2000, ISBN 0465007805.