OT: cheap colo space in Southern Germany/Munich

Scott Weeks surfer at mauigateway.com
Mon Nov 26 21:37:36 CET 2012

Sounds like a lot of work for something that's not necessary to the
operation of the network.  SNSH (Stupid Network - Smart Hosts :)

>> What, it's not Friday yet? Unsend, unsend, unsend.


--- eugen at leitl.org wrote:

From: Eugen Leitl <eugen at leitl.org>
To: ipv6-ops at lists.cluenet.de
Subject: Re: OT: cheap colo space in Southern Germany/Munich
Date: Mon, 26 Nov 2012 21:27:00 +0100

On Mon, Nov 26, 2012 at 11:56:26AM -0800, Scott Weeks wrote:

> It's still earth-centric.  As I said above, why impose an 
> artificial boundary?  Remove any bounds possible on the
> network, so it can expand as necessary.

It's not Earth centric, you can subnet all noticable
gravity wells in the solar system, or the supercluster,
for that matter. E.g. even a /64 is enough for WGS84
for Earth, with 24 bits for latitude and longitude,
and 16 bit for height (+/- 32 km ought to be enough
for everybody ;). Resolution is about a node/m^3 at
Earth surface. More than enough for wireless nodes,
especially if you allow for local distortions for
node assemblies (jiggle the LSB randomly until
uncollided). Add DTN
and consider that there are experimental Android-phone
driven low LEO sats, add 4G, or realtime phased array
beamforming for flyby retargeting (of course the phone
has all the ephemerides, and knows it's realtime
orientation in space even if it's tumbing, or uses 3 
orthogonally aligned coils for steering in Earth's 
magnetic field), and it's going to get quite interesting.

Notice that that would be routable across the IPv6
Internet, as the local /64 can be chosen arbitrarily, and can
be even made RFC-compliant.

I agree that IPv6 has insufficient bits, if you want
to be able drill down to a cubic micron, or even nm^3
anywhere in the local solar system.

So even IPv6 will have legs its designers never
anticipated (if they did, they'd pick at least 256 bit).
Due to such novel uses IPv6 exhaustion will probably
happen much sooner than most people think.

I have the impression that free space LoS
in GBit/s to TBit/s in space will come pretty
soon now. There are multiple mentions of diverse 
projects, some vapor, some (apparently) solid


Notice that LoS across air or vacuum has
a far lower relativistic latency, unless a
photonic crystal waveguide can be designed which
would allow near-c speeds in fiber.

Most hardware will be in space, eventually.
Cloud must be in the sky, where it belongs ;)

What, it's not Friday yet? Unsend, unsend, unsend.

More information about the ipv6-ops mailing list