From: Eugene Leitl (eugene@liposome.genebee.msu.su)
Date: Tue Jul 07 1998 - 08:39:02 MDT
Dwayne writes:
> This is true. When I brought up the PunkNet idea
> (http://i.am/punknet) on the cypherpunks list some years ago,
Interesting, thanks for the URL. Unfortunately can't access it just
now as darned local DNS is down, again :(
> there was quite a bit of discussion regarding how to set up an
> unbreakable communications network over intercontinental
> distances. There weren't a lot of decent responses, as the
> problem is non-trivial.
In the first iteration, you can simply use existing communication matrix:
cellular, two-way pagers, wire telephony and the Internet (soon from
the skies?), overlaying them with a crypted higher-order protocol
(heck, even using ssh or its free GNU clone would suffice just now).
Afaik encryption is not tolerated in HAM circles, so using existing
packet radio infrastructure won't be possible. Of course things like
that will be down in course of a true calamity, so a minimal backup
capability will be certainly necessary. (This is always assuming the
calamity is serious, but transient. In case of quantitative holocaust
there is obviously little need to maintain connectivity).
> But local nets, yes, are a good idea. I'm working on doing
> something on a metropolitan scale across Melbourne, Australia,
> I'll let anyone interested know how it pans out.
You talked about spread-spectrum cellular, which sounds very interesting.
Quite a time ago on this list networking-relevant results of Stu
Kauffman's work were mentioned: there is a minimal, low connectivity beyond
which a randomly connected net experiences a phase transition into a
single contiguous domain. PV-driven (supercapacitor-buffered?) cheap
spread-spectrum cellulars (think about NEMP when designing these!)
capable of routing TCP/IP(ng) should be of considerable interest
to the mainstream wireheads as cheap, low-bandwidth/long-latency
Internet infrastructure. If enough enthusiasts will be operating these
(difficult to make such a movement supercritical: you need a
cheap available kit and a minimal density to make the coverage
useful for anything), you'll have an alternative communication
infrastructure. Microwave and line-of-sight diode laser links
could be also useful.
Albeit this is ridiculous to contemplate just now, lauching your very own
LOSL palmtop relays into LEO should become not too expensive in a few decades.
> [...]
> I think that looking at the history of the last 200 years will
> give anyone rational food for thought.
Right. For starters, getting germtight suits into mass production and
into everybody's living room and making sure the majority can use them
(remember the suffocated children in Israel during the Gulf war?)
would mean a lot (well, a great deal more than a working protection
against random asteroid impact). They would also offer excellent
protection against fallout (how many have sturdy basements, a
supply of clean water, dehydrated food and a basic medical kit,
atropine, a dosimeter and iodide tablets included?), and limited
protection against C weapons. Of course this is semiuseless against
a sneaky attack with long-latency engineered bioweapons (which will
need sniffers, and continuous population screeing on viral DNA
kinetics), and certainly useless against a Gray Goo incident as
in case of a (spontaneous, or engineered) Blight.
In any case will the effort not be wasted, just think of MDR epidemics
and reactors gone China.
[...]
> Um. Why is it possible to find this list in hotbot? Why isn't it
> closed to search engines?
This is meant as an open discussion/education forum. A fashistoid
robot.txt would render the archives websearch-invisible, which
would seem to defy their purpose.
> I find the prospect of having my postings readable by the whole
> internet quite disturbing, as I may well feel free to say some
> things here which I might not want to be known about by the
> general public.
Well, start a list which distributes only signed, crypted
stuff (is there a hacked majordomo which does this?), and
is open only to known members. You're virtually certain to draw
interest from the wrong type of people, then, security included.
The latter has started to be pretty paranoid since the Aum
et al. type of thing.
> Dwayne
'gene
This archive was generated by hypermail 2.1.5 : Fri Nov 01 2002 - 14:49:18 MST