... circumvention.¹

Davis and Geer point out in [DGT96] that, in some contexts such as Kerberos, challenge-response can take the place of accurate timestamps.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... port.²

This was accomplished using the author's parallel port pin programming library for Linux, parapin, which is freely available at http://www.circlemud.org/jelson/software/parapin

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...post-facto. ³

``The joy of engineering is to find a straight line on a double logarithmic diagram.'' -Thomas Koenig

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... sensors⁴

In this experiment, we assume that each sensor is preprogrammed with its position in some global coordinate system. In the context of a testbed meant to demonstrate time synchronization, this assumption is a reasonable one. In parallel, Girod is developing techniques that allow a distributed federation of sensors to dynamically create a self-consistent coordinate system.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.