Endgame: Creation of DDI via radical innovation

3 views
Skip to first unread message

solarnetone

unread,
Feb 28, 2011, 4:00:18 AM2/28/11
to Building a Distributed Decentralized Internet, sc...@solarnetone.org
Hi All,

Having read some profiles and posts, I am going to throw out a few
ideas based on the premise that "if you do not like the management
methods of the current system, make your own system." The system in
question is the lowest level of the network. We can power our
machines autonomously via a variety of means. We can make our
machines run stably and relatively securely. What we cannot control,
and what seems to be the point of the discussion, is the upstream
network's stability, politics, and barriers to access. The point here
is that we should not throw out the baby with the bathwater simply
because satellite and cellular data service is expensive, cable and
phone lines are filtered, and mesh networks are not yet robust or
tuned well enough for the liking of most. By addressing the problem
at the correct layer, it can be most easily solved.

OK, time for some raw ideas for a new physical layer not yet in
existence... get there first, and its yours to do with as you please,
being the gist of the idea:
TCP/IP over quantum entanglement
TCP/IP over frequency ranges not yet used... http://www.blazelabs.com/pics/em_spectplanck.gif
TCP/IP over Tesla's scalar impulses

All require a roomful of specialized genius type folks. Personally, I
think the latter of the above will be the easiest to bring to fruition
and will most closely resemble the desired final outcome, although I
have thoughts on all three, if anyone is willing to discuss pushing
back the edge of reality a bit to meet the needs of civilization.

Cheers,
Scott

Robert Steele

unread,
Feb 28, 2011, 8:01:39 AM2/28/11
to building-a-distributed...@googlegroups.com
very cool.  see Free Space Optics, I understand this is on the verge of breaking wide open.

Urge everyone to start migrating their own google group contributions to the wiki, using Michel's protocol of naming the person making the statement, and my original suggestion, linking back to the original googlegroup email for sourcing.

When DEMO is done, I will start doing what I can on my own, working from the bottom up.

Eugen Leitl

unread,
Feb 28, 2011, 8:10:00 AM2/28/11
to building-a-distributed...@googlegroups.com
On Mon, Feb 28, 2011 at 08:01:39AM -0500, Robert Steele wrote:
> very cool. see Free Space Optics, I understand this is on the verge of
> breaking wide open.

This isn't new (Ronja has been around forever), and it's extremely
sensitive to alignment, and atmospheric condition. In general,
we have 2.4, 5 and soon 60 GHz for long-range WiFi, which are
comparable in reach when used with directional aerials while
not as sensitive to atmospheric state and alignment.



> Urge everyone to start migrating their own google group contributions to the
> wiki, using Michel's protocol of naming the person making the statement, and
> my original suggestion, linking back to the original googlegroup email for
> sourcing.
>
> When DEMO is done, I will start doing what I can on my own, working from the
> bottom up.

--
Eugen* Leitl <a href="http://leitl.org">leitl</a> http://leitl.org
______________________________________________________________
ICBM: 48.07100, 11.36820 http://www.ativel.com http://postbiota.org
8B29F6BE: 099D 78BA 2FD3 B014 B08A 7779 75B0 2443 8B29 F6BE

James Valleroy

unread,
Mar 4, 2011, 11:54:15 AM3/4/11
to The Next Net
Has anyone considered ultrasound? Possibly with directional
speakers / microphones. I don't know what kind of range it would get
though.

On Feb 28, 4:00 am, solarnetone <twinbeardthepir...@gmail.com> wrote:
> Hi All,
>
> Having read some profiles and posts, I am going to throw out a few
> ideas based on the premise that "if you do not like the management
> methods of the current system, make your own system."  The system in
> question is the lowest level of the network.  We can power our
> machines autonomously via a variety of means.  We can make our
> machines run stably and relatively securely.  What we cannot control,
> and what seems to be the point of the discussion, is the upstream
> network's stability, politics, and barriers to access.  The point here
> is that we should not throw out the baby with the bathwater simply
> because satellite and cellular data service is expensive, cable and
> phone lines are filtered, and mesh networks are not yet robust or
> tuned well enough for the liking of most.  By addressing the problem
> at the correct layer, it can be most easily solved.
>
> OK, time for some raw ideas for a new physical layer not yet in
> existence... get there first, and its yours to do with as you please,
> being the gist of the idea:
> TCP/IP over quantum entanglement
> TCP/IP over frequency ranges not yet used...http://www.blazelabs.com/pics/em_spectplanck.gif

Aaron Huslage

unread,
Mar 4, 2011, 12:43:35 PM3/4/11
to building-a-distributed...@googlegroups.com

Too low a frequency. Data rates would be in the single digit bps I would imagine...dtn or not that's too slow. Also it would have to be REALLY loud.

--
Aaron Huslage
+1.919.600.1712
hus...@gmail.com
Via mobile device

solarnetone

unread,
Mar 4, 2011, 2:09:05 PM3/4/11
to The Next Net
I was thinking along the lines of Petahertz bands...

On Mar 4, 12:43 pm, Aaron Huslage <husl...@gmail.com> wrote:
> Too low a frequency. Data rates would be in the single digit bps I would
> imagine...dtn or not that's too slow. Also it would have to be REALLY loud.
>
> --
> Aaron Huslage
> +1.919.600.1712
> husl...@gmail.com
> Via mobile device
Reply all
Reply to author
Forward
0 new messages