On Fri, 16 Mar 2001, Damien Broderick wrote:
> At 08:37 PM 15/03/01 -0800, Robert wrote:
>
> >you can assign a real "cost" to losing energy
> >(due to photon spreading due the inverse square law affecting
> >the strength of signals being transmitted -- photons you can't
> >receive with your "detector" are lost energy/matter) and the actual
> >"matter" lost due to ablation
>
> Redundancy. Redundancy. Redundancy. Let me say it again. Redundancy.
>
> Cost goes up, true.
>
It isn't the error rate I'm harping about its the "lost" energy {=matter}.
Spike and I have had this discussion off-list. Theoretically at least,
if you send photons acros a large distance you can "regenerate" the
energy used to send them (the principle behind CCD's is that the
photons knock an electron into the conduction band and that electron
gets used to generate a "charge" that you could later use downstream.
Similarly, you accelerate a DTU (data transmission unit) up to 0.1c
in a mass driver, then decelerate it in a similar unit after it
travels N light years and you recapture the kinetic energy it contains
(at whatever your conversion efficiencies are).
But both transmission strategies have losses -- the EM wave transmission
strategy loses photons unless you have a mirror large enough to capture
them all (and that has to be really big at 1 light year out...). The
DTU loses mass to erosion by interstellar gas and dust.
However the kicker is that you can control how much mass you lose in
the interstellar DTU by altering its velocity. You can't control
the losses due to the spreading of your electromagnetic communication
beam because the dispersion is controlled by the laws of physics.
In a fully developed solar system, you are short on both *energy*
and *matter* -- so if you are going to communicate or send probes
(or colonizing expediations), you better make sure you do it
as efficiently as possible and get back more than you have
invested. Because of the inherent physical losses in EM waves
and their low information carrying capacity (relative to
a couple of kg of matter), DTU's seem to be highly preferred.
The only exceptions would appear to be if you can think of
situations in which you can "trade" ~10^-20 to ~10^-30 of
the information you already have at your disposal over a link
that is relatively expensive and get "perceived" value from
that information that comes back that justifies the cost. I haven't
done the translation into how many bits of "new" information
you can generate locally relative to what you can receive
from a remote exhange at a constant cost (in part this
depends on the "novelty" of the remotely generated vs.
the locally generated information and the distance over which
the local<->remote exchange is occuring -- but I suspect the
remote information is going to have to be really novel
to justify the expense.
Robert
This archive was generated by hypermail 2b30 : Mon May 28 2001 - 09:59:40 MDT