# [FoRK] FoRK Digest, Vol 89, Issue 10

Dave Long dave.long at bluewin.ch
Sun Feb 13 10:17:41 PST 2011

```> At 10 GBit, a bit is roughly 3 cm in vacuum, slightly shorter in
> glass.

Olympic fencing is a twitch game.  For instance, in order to count a
clean hit in épée, the scoring touch has to land at least 40 ms
before the opponent's point.

Given that human choice reaction times are far longer, on the order
of 150-200 ms[0], one might think that double hits would only occur
in the rare cases where the opponents coincidentally decided to
attack at the same time.

However, double hits are not that uncommon, and my current
explanation for this apparent paradox is that closing the distance to
make an attack, even without the defender's conscious reaction,
symmetrically lessens the distance that the defender's point must
travel before touching the attacker; therefore we should expect just
about any attack launched from a neutral position to result in a
double hit.  (hence the relative importance of preparing an attack by

The worst case occurs when the putative defender takes advantage of
the collapse of distance to launch a counterattack, thereby
collapsing it even faster.  If we model the point motion in a simple
attack as closing from 0 to 4 m/s to 0 over about 600 ms, and the
defender's counterattack starts when they first react, at 150 ms past
the initial attack, there will be a significant plateau where both
points are moving towards target at 6 m/s [1].

((6 (m / s)) * 40 ms) / (3 (cm / bit)) = 1 octet

Therefore (albeit entirely coincidentally) the advantage in position
an épéiste should maintain to get a clean hit is (at 10GB) about a byte.

-Dave

[0] it is no coincidence that the notion of a "fencing time" within a
phrase is on the order of 140-180 ms.
[1] this explains the difference between cavalry sabers and fencing
swords; two cavaliers at full tilt close on each other much faster,
with roughly 24 m/s.  For such an engagement the "fencing time" is
about 3,6m, well over a horse length.

```