Defining the problem

40 views
Skip to first unread message

Colin Hawkett

unread,
Feb 25, 2011, 1:47:36 PM2/25/11
to building-a-distributed...@googlegroups.com
This might not be the most popular perspective, but I'm not sure everybody is trying to solve the same problem here, or for the same reasons. There's some complicated socio-political agendas and different people see different stuff - the path from A->B is assumed obvious - indeed that A and B are obvious. It might just be me who's not sure :)

Is it that we want to be able to use the web we know today on a decentralised hardware and software infrastructure? Why? Which parts of that problem are we proposing to solve? By what path? To what extent? For what benefit?

Decentralised is an ambiguous terminology. e.g. is a local community wireless transmitter centralised or decentralised?

Basically what's the problem, what are the options, what are their advantages/disadvantages, on what time scales will they manifest, under what constraints & assumptions. What are the outcomes we are looking for, and can we measure success/failure?

A distributed internet isn't an outcome - it facilitates an outcome. What is that?

Samuel Rose

unread,
Feb 25, 2011, 2:11:20 PM2/25/11
to building-a-distributed...@googlegroups.com, Colin Hawkett
For myself, personally:

On Fri, Feb 25, 2011 at 1:47 PM, Colin Hawkett <haw...@gmail.com> wrote:
> This might not be the most popular perspective, but I'm not sure everybody
> is trying to solve the same problem here, or for the same reasons. There's
> some complicated socio-political agendas and different people see different
> stuff - the path from A->B is assumed obvious - indeed that A and B are
> obvious. It might just be me who's not sure :)
> Is it that we want to be able to use the web we know today on a
> decentralised hardware and software infrastructure? Why?

Yes, this is *part* of it. We should be able to take existing open
protocols and web technologies and run them on a distributed internet.
Why? It doesn't make sense to me to try and leave that behind right
now, plus it can work with what will very likely be a reasnable
investment on targetted development in the traditional web application
"stack". Some emerging web applications may work right out of the box
(couchdb etc etc)

> Which parts of that
> problem are we proposing to solve? By what path? To what extent? For what
> benefit?

In the context of contactcon, we want to keep the floor open, we want
to support the various problems people are trying to solve in
parallel. I don't want to just focus in on one problem, because that
will exclude others that I think are important. There are a plurality
of problems. That might not seem helpful for focusing in on "what can
be done now specifically" but let's bring some more players to the
discussion, and more people working on more parts of the picture
before we try to narrow down the focus, is what I personally think.
We're in the process of talking directly with people working on the
hardware, protocol and software layers now.

I think the next stage is to get those folks here to answer all of
these questions we're asking. Then, when there more stakeholders, we
can draw some boundaries which will narrow things down. Just my
opinion. I think we need at least another week or 2 before we start
focusing in/narrowing down, if only to give us a chance to get more
stakeholders in the discussion. I am open minded though, and this is
only my personal take on it.

> Decentralised is an ambiguous terminology. e.g. is a local community
> wireless transmitter centralised or decentralised?
> Basically what's the problem, what are the options, what are their
> advantages/disadvantages, on what time scales will they manifest, under what
> constraints & assumptions. What are the outcomes we are looking for, and can
> we measure success/failure?
> A distributed internet isn't an outcome - it facilitates an outcome. What is
> that?
>

I know many people who probably would disagree that a distributed
internet is not an outcome. Although they might agree that it
facilitates many outcomes. I think we've documented thousands of those
outcomes that a distributed internet facilitates at
http://p2pfoundation.net throughout that wiki over the last 5 years.

Still, I understand your frustration, and reasoning for wanting to
focus in. Personally, I'd like to see a distributed hardware/network
medium where people can pool hardware resources and network connection
node resources to run some variant of practically anything they can do
on the internet now (starting with the simplest things, like message
passing, etc).


--
--
Sam Rose
Future Forward Institute and Forward Foundation
Tel:+1(517) 639-1552
Cel: +1-(517)-974-6451
skype: samuelrose
email: samue...@gmail.com
http://forwardfound.org
http://futureforwardinstitute.org
http://hollymeadcapital.com
http://p2pfoundation.net
http://socialmediaclassroom.com

"The universe is not required to be in perfect harmony with human
ambition." - Carl Sagan

Samuel Rose

unread,
Feb 25, 2011, 2:53:43 PM2/25/11
to building-a-distributed...@googlegroups.com, Colin Hawkett
Ps. The reply below is definitely not meant to dismiss what you are
talking about, Colin. Because I think we need to do exactly what you
are suggesting, which is to focus in on specifics. For my part, I'd
just like to try and get some more people here and get them all
involved in focusing in together with us.

Robert Steele

unread,
Feb 25, 2011, 3:12:06 PM2/25/11
to building-a-distributed...@googlegroups.com, co...@cookreport.com, da...@mindtel.com
At the link below is Arron's contribution, which strikes me as having enormous potential, followed by a much longer and more difficult to digest report from Gordon Cook (tip of the hat to Venessa). 

Reference: Aidphone Flybox for Autonomous Internet

Reference: Cook Report on Internet Protocol V4 E Pluribus Unum* Resurrected: How Human Ingenuity, DIY Technology, and Global R&E Networks Are Remaking the World Part 1 of 2 Parts


I have started my article/chapter on

Open Everything:

From Autonomous Internet to Global Panarchy


and what I am looking for and not seeing is a table of applicable open source softwares, hardwares, and affordable satellite access leased access.  Dave Warner and a few others did some really cool stuff with STRONG ANGEL and TOOZL, and the DARPA moved on to more escoteric less useful stuff.

Am not a technical person, very interested in getting to a very detailed table of solutions that can be broadcast.

Robert

Paul B. Hartzog

unread,
Feb 25, 2011, 4:54:22 PM2/25/11
to building-a-distributed...@googlegroups.com, Robert Steele, co...@cookreport.com, da...@mindtel.com
I have great diagrams for how to visualize these things,
but I need to get in a room with a graphic designer for a few hours
to produce some good visualizations.

Any ideas?

-p

--
--------------------------------------------------------
The Forward Foundation
http://www.ForwardFound.org
paul.b....@forwardfound.org
--------------------------------------------------------
http://www.PaulBHartzog.org
PaulBH...@PaulBHartzog.org
--------------------------------------------------------
http://www.panarchy.com
PaulBH...@panarchy.com
--------------------------------------------------------
University of Michigan
PHar...@umich.edu
--------------------------------------------------------
The Universe is made up of stories, not atoms.
                 --Muriel Rukeyser

Perceive differently, then you will act differently.
                 --Paul B. Hartzog
--------------------------------------------------------

Robert Steele

unread,
Feb 25, 2011, 4:58:13 PM2/25/11
to paulbh...@gmail.com, building-a-distributed...@googlegroups.com, co...@cookreport.com, da...@mindtel.com
I have a great graphic designer who charges 35 an hour--does not get better than that.  I also have a young adult at the Rochester Institute of Technology.

Attached is my really ugly start on the paper.  I hope it shows what I want to do.  I would like to find a way to integrate everything everyone here knows about creating an autonomous internet into this paper, and make it something that goes everywhere and gets translated.

This paper, which certainly could and should be co-authored, is the perfect place to put in graphics, and once the paper is completed, then it reverts to the wiki (or the other way around, but at my age it is easier to do it as an article first).
Open Everything Draft 1.0.doc

Devin Balkind

unread,
Feb 25, 2011, 4:57:25 PM2/25/11
to building-a-distributed...@googlegroups.com
Where does this room need to be?

DWB

On Feb 25, 2011, at 16:54, "Paul B. Hartzog" <paulbh...@gmail.com>
wrote:

> I have great diagrams for how to visualize these things,

Samuel Rose

unread,
Feb 25, 2011, 5:19:35 PM2/25/11
to building-a-distributed...@googlegroups.com, Paul B. Hartzog, Robert Steele, co...@cookreport.com, da...@mindtel.com
Maybe we should at least sketch up first using something like
inkscape? I'd be glad to spend the time sometime really soon

--

Eugen Leitl

unread,
Feb 25, 2011, 5:41:09 PM2/25/11
to building-a-distributed...@googlegroups.com
On Fri, Feb 25, 2011 at 05:19:35PM -0500, Samuel Rose wrote:
> Maybe we should at least sketch up first using something like
> inkscape? I'd be glad to spend the time sometime really soon

Please do not top-post and please trim your mail.

Just as an aside: you soft humanities people are failing already
even before you started. This is not how it works.

Can you read RFCs? Can you write RFCs? Do you at least
know people who will do it to your specs?

--
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

Venessa Miemis

unread,
Feb 25, 2011, 6:02:50 PM2/25/11
to Building a Distributed Decentralized Internet
hey eugen!

sorry if i haven't put a place on this google group yet that explains
the intentions..... i'm organizing a gathering in nyc in october with
400 of the top people i can find in the world who are working on the
various aspects of the emerging networked society. i started this
group to get some conversation and action going. one of the
initiatives that is coming together here is to talk about distributed
internet infrastructures. because we are promoting open discussion and
collaborative solutions, this is a place where the big picture takes
precedent over distinctions over theory verse practice. i think this
movement as a whole has been stunted for years because it's so
splintered. every faction seems to want to claim that its aspect is
the essential one. i think it's worth a try to be inclusionary and
understand that a problem of this scale and complexity is going to
need A LOT of intelligent brains working together. i think it's gonna
be tough at first to figure out how to find some common language
between the builders and the thinkers and all the various people
involved, but i think we can.

please contribute your voice and your opinions and your community and
your work... but please don't prevent others to bring their ideas,
concerns, and projects to the table as well.

i think we have a chance here to see how all of our skills and
knowledge can fit together in a bigger way.

- venessa
> > > paul.b.hart...@forwardfound.org
> > > --------------------------------------------------------
> > >http://www.PaulBHartzog.org
> > > PaulBHart...@PaulBHartzog.org
> > > --------------------------------------------------------
> > >http://www.panarchy.com
> > > PaulBHart...@panarchy.com
> > > --------------------------------------------------------
> > > University of Michigan
> > > PHart...@umich.edu
> > > --------------------------------------------------------
> > > The Universe is made up of stories, not atoms.
> > > � � � � � � � �� --Muriel Rukeyser
>
> > > Perceive differently, then you will act differently.
> > > � � � � � � � �� --Paul B. Hartzog
> > > --------------------------------------------------------
>
> > --
> > --
> > Sam Rose
> > Future Forward Institute and Forward Foundation
> > Tel:+1(517) 639-1552
> > Cel: +1-(517)-974-6451
> > skype: samuelrose
> > email: samuel.r...@gmail.com
> >http://forwardfound.org
> >http://futureforwardinstitute.org
> >http://hollymeadcapital.com
> >http://p2pfoundation.net
> >http://socialmediaclassroom.com
>
> > "The universe is not required to be in perfect harmony with human
> > ambition." - Carl Sagan
>
> --
> Eugen* Leitl <a href="http://leitl.org">leitl</a>http://leitl.org
> ______________________________________________________________
> ICBM: 48.07100, 11.36820http://www.ativel.comhttp://postbiota.org

Robert Steele

unread,
Feb 25, 2011, 6:03:22 PM2/25/11
to building-a-distributed...@googlegroups.com
Okay, before we have a fight, here is the Ref A graphic from us humanities people,
and Ref B big picture.

Graphic: Web of Fragmented Knowledge

2010 M4IS2 Briefing for South America — 2010 M4IS2 Presentacion por Sur America (ANEPE Chile)

I am going to journey to New Jersey on or about 4 March to debriefing Gordon Cook,
am wondering if a round-table someplace convenient for all of us, in late March, would
be cool.  Or a Go to Meeting thing where you technical people talk and we humanities
people comment with wit and wisdom?

I hope Aaron is on this list, I think you guys have so much to offer, and that it needs to be
structured and harvested sooner than later.

Connexus Sumus Unum....
Robert

Venessa Miemis

unread,
Feb 25, 2011, 6:08:07 PM2/25/11
to Building a Distributed Decentralized Internet
good call robert... there are still a bunch of really smart folks who
i'm only communicating with via email... i'll send them invites here
now.

- v

On Feb 25, 6:03 pm, Robert Steele
<robert.david.steele.vi...@gmail.com> wrote:
> Okay, before we have a fight, here is the Ref A graphic from us humanities
> people,
> and Ref B big picture.
>
> Graphic: Web of Fragmented
> Knowledge<http://www.phibetaiota.net/2010/09/2008/08/graphic-web-of-fragmented-...>
>
> 2010 M4IS2 Briefing for South America — 2010 M4IS2 Presentacion por Sur
> America (ANEPE Chile)<http://www.phibetaiota.net/2010/12/2010-m4is2-briefing-for-south-amer...>
>
> I am going to journey to New Jersey on or about 4 March to debriefing Gordon
> Cook,
> am wondering if a round-table someplace convenient for all of us, in late
> March, would
> be cool.  Or a Go to Meeting thing where you technical people talk and we
> humanities
> people comment with wit and wisdom?
>
> I hope Aaron is on this list, I think you guys have so much to offer, and
> that it needs to be
> structured and harvested sooner than later.
>
> Connexus Sumus Unum....
> Robert
>
>
>
> On Fri, Feb 25, 2011 at 5:41 PM, Eugen Leitl <eu...@leitl.org> wrote:
> > On Fri, Feb 25, 2011 at 05:19:35PM -0500, Samuel Rose wrote:
> > > Maybe we should at least sketch up first using something like
> > > inkscape? I'd be glad to spend the time sometime really soon
>
> > Please do not top-post and please trim your mail.
>
> > Just as an aside: you soft humanities people are failing already
> > even before you started. This is not how it works.
>
> > Can you read RFCs? Can you write RFCs? Do you at least
> > know people who will do it to your specs?
>
> > > On Fri, Feb 25, 2011 at 4:54 PM, Paul B. Hartzog <paulbhart...@gmail.com>
> > > > paul.b.hart...@forwardfound.org
> > > > --------------------------------------------------------
> > > >http://www.PaulBHartzog.org
> > > > PaulBHart...@PaulBHartzog.org
> > > > --------------------------------------------------------
> > > >http://www.panarchy.com
> > > > PaulBHart...@panarchy.com
> > > > --------------------------------------------------------
> > > > University of Michigan
> > > > PHart...@umich.edu
> > > > --------------------------------------------------------
> > > > The Universe is made up of stories, not atoms.
> > > >                  --Muriel Rukeyser
>
> > > > Perceive differently, then you will act differently.
> > > >                  --Paul B. Hartzog
> > > > --------------------------------------------------------
>
> > > --
> > > --
> > > Sam Rose
> > > Future Forward Institute and Forward Foundation
> > > Tel:+1(517) 639-1552
> > > Cel: +1-(517)-974-6451
> > > skype: samuelrose
> > > email: samuel.r...@gmail.com
> > >http://forwardfound.org
> > >http://futureforwardinstitute.org
> > >http://hollymeadcapital.com
> > >http://p2pfoundation.net
> > >http://socialmediaclassroom.com
>
> > > "The universe is not required to be in perfect harmony with human
> > > ambition." - Carl Sagan
> > --
> > Eugen* Leitl <a href="http://leitl.org">leitl</a>http://leitl.org
> > ______________________________________________________________
> > ICBM: 48.07100, 11.36820http://www.ativel.comhttp://postbiota.org

Samuel Rose

unread,
Feb 25, 2011, 6:11:45 PM2/25/11
to building-a-distributed...@googlegroups.com, Eugen Leitl
On Fri, Feb 25, 2011 at 5:41 PM, Eugen Leitl <eu...@leitl.org> wrote:
> On Fri, Feb 25, 2011 at 05:19:35PM -0500, Samuel Rose wrote:
>> Maybe we should at least sketch up first using something like
>> inkscape? I'd be glad to spend the time sometime really soon
>
> Please do not top-post and please trim your mail.
>
> Just as an aside: you soft humanities people are failing already
> even before you started. This is not how it works.
>
> Can you read RFCs? Can you write RFCs? Do you at least
> know people who will do it to your specs?
>

Eugen, fair enough on the top-post/trimming.

Paul B. Hartzog, Richard C. Adler and myself have written an RFC.
We're not "soft humanities people".

I don't think we are at the stage in *this* discussion to start
writing RFC's, personally. Although I do agree that writing RFC(s)
could be an early activity that we could take on doing very soon.

Venessa Miemis

unread,
Feb 25, 2011, 6:15:29 PM2/25/11
to Building a Distributed Decentralized Internet
eugen,

again, this google group only got started 2 or 3 days ago. we are just
starting to contact the communities of people who are interested in
collaborating. please give me a chance to get this going! it sounds
like you're a knowledgeable programmer, so i appreciate you being
here. right now we have a mix of philosophers and programmers here.
please give it a chance to see that we can coexist in this
conversation.

the skills that you have that others don't is valuable, as is the
knowledge they have that you don't. no one here is trying to be right.
we're just trying to understand the bigger picture and where the
opportunities are to build, collaborate, and understand what is really
going on here.

literally, on a global scale right now, people are paying attention to
this, and they think it matters. now is the time to figure out how we
can cooperate so everybody moves forward.

- v

On Feb 25, 5:41 pm, Eugen Leitl <eu...@leitl.org> wrote:
> > > paul.b.hart...@forwardfound.org
> > > --------------------------------------------------------
> > >http://www.PaulBHartzog.org
> > > PaulBHart...@PaulBHartzog.org
> > > --------------------------------------------------------
> > >http://www.panarchy.com
> > > PaulBHart...@panarchy.com
> > > --------------------------------------------------------
> > > University of Michigan
> > > PHart...@umich.edu
> > > --------------------------------------------------------
> > > The Universe is made up of stories, not atoms.
> > > � � � � � � � �� --Muriel Rukeyser
>
> > > Perceive differently, then you will act differently.
> > > � � � � � � � �� --Paul B. Hartzog
> > > --------------------------------------------------------
>
> > --
> > --
> > Sam Rose
> > Future Forward Institute and Forward Foundation
> > Tel:+1(517) 639-1552
> > Cel: +1-(517)-974-6451
> > skype: samuelrose
> > email: samuel.r...@gmail.com
> >http://forwardfound.org
> >http://futureforwardinstitute.org
> >http://hollymeadcapital.com
> >http://p2pfoundation.net
> >http://socialmediaclassroom.com
>
> > "The universe is not required to be in perfect harmony with human
> > ambition." - Carl Sagan
>
> --
> Eugen* Leitl <a href="http://leitl.org">leitl</a>http://leitl.org
> ______________________________________________________________
> ICBM: 48.07100, 11.36820http://www.ativel.comhttp://postbiota.org

Devin Balkind

unread,
Feb 25, 2011, 6:51:13 PM2/25/11
to building-a-distributed...@googlegroups.com
Eugen. Enthusiastic students do amazing work so if you'd like to talk
technical, please post your documentation and we will start following
your rss at floing and provide you support.

If someone doesn't start a Tiddlywiki for this group before tomorrow I
will start one at tiddlyspot.

The other things people have posted on this board have been very
exciting and look forward to posting more when I get off the road
tomorrow.

DWB

On Feb 25, 2011 at 18:08, Venessa Miemis <veness...@gmail.com>
wrote:

> good call robert... there are still a bunch of really smart folks who
> i'm only communicating with via email... i'll send them invites here
> now.
>

> - v
>

>>>>> --Muriel Rukeyser
>>
>>>>> Perceive differently, then you will act differently.

Robert Steele

unread,
Feb 25, 2011, 6:57:31 PM2/25/11
to building-a-distributed...@googlegroups.com
I second Venessa's comment.  We have an opportunity in the next few weeks to
get some serious attention to options for building an autonomous internet and
spreading that knowledge.  Egypt was a catalyst, and my Internet Work-Arounds,
which as I recall got huge extra stuff from one of you, is one of the top hits for the
past several weeks. 

My intent with the article is to create a roadmap toward open everything, not just
a list, but an executable road map starting with the open tri-fecta--software,
spectrum, and open source intelligence (decision-support and deliberation tools).

I see satellites, and nano-satellites, and WiMax, and pigions, all coming into play.
"Whatever it takes" to include breaking spectrum rules that are antiquated, when
necessary for in extremis support to good causes.

The US Government is so hosed on spectrum they cannot run half the stuff they
have in Afghanistan without spectrum conflicts.  This is a good time for a
breakout, IMHO.

Samuel Rose

unread,
Feb 25, 2011, 8:07:51 PM2/25/11
to building-a-distributed...@googlegroups.com, Devin Balkind
On Fri, Feb 25, 2011 at 6:51 PM, Devin Balkind <devinb...@gmail.com> wrote:
> Eugen.  Enthusiastic students do amazing work so if you'd like to talk
> technical, please post your documentation and we will start following your
> rss at floing and provide you support.
>
> If someone doesn't start a Tiddlywiki for this group before tomorrow I will
> start one at tiddlyspot.
>
> The other things people have posted on this board have been very exciting
> and look forward to posting more when I get off the road tomorrow.
>
> DWB
>

Devin, I need to look at Tiddlywiki again, it's been a long time. I
used to use it as a Desktop wiki a while ago, and my friend Hans Wobbe
uses it extensively in his business. I need to re-investigate it
because if it is still the same technology that it was 5-6 years ago,
it is an interesting candidate for "distributed internet" web
application. It'd be interesting for instance to extend TW with
javascript, think about storage/archiving with couchdb (which itself
can be redundantly distributed as a datastore) etc.


--
--
Sam Rose
Future Forward Institute and Forward Foundation
Tel:+1(517) 639-1552
Cel: +1-(517)-974-6451
skype: samuelrose

email: samue...@gmail.com

Samuel Rose

unread,
Feb 25, 2011, 8:12:34 PM2/25/11
to building-a-distributed...@googlegroups.com, Robert Steele

Is there anyone present that is knowledgeable about the existing US
rules (and EU rules too) around radio spectrum? Particularly
http://en.wikipedia.org/wiki/White_spaces_(radio) in US?

Reading through
http://hraunfoss.fcc.gov/edocs_public/attachmatch/FCC-10-174A1.pdf and
if anyone else knows what is possible now for packet forwarding, using
radio relays in unlicensed spectrum etc please share here. Thanks!


--
--
Sam Rose
Future Forward Institute and Forward Foundation
Tel:+1(517) 639-1552
Cel: +1-(517)-974-6451
skype: samuelrose

email: samue...@gmail.com

Aaron Huslage

unread,
Feb 25, 2011, 8:24:15 PM2/25/11
to building-a-distributed...@googlegroups.com
Raises hand!

My friend Sheldon Renan also has a lot to say on these topics. I've invited him to come play. He's got another concept in the Cook Report called Netness which is insanely compelling (I'm biased, I helped him with some of the technical aspects and was an early believer). I can't find the URL right now, but I'll post it when I can.

Aaron Huslage

unread,
Feb 25, 2011, 8:36:38 PM2/25/11
to building-a-distributed...@googlegroups.com, Samuel Rose, Robert Steele
Is there anyone present that is knowledgeable about the existing US
rules (and EU rules too) around radio spectrum? Particularly
http://en.wikipedia.org/wiki/White_spaces_(radio)  in US?

EU has no white spaces yet, as far as I'm aware. The US recently announced its rules so white spaces protocol finalization and use should begin in the next year or two...if not sooner.

In general, the FCC rules are pretty clear. There are Part-15 bands (also called ISM) where any unlicensed user can transmit up to a ceiling of power (100mW for 2.4gHz, etc). These devices must receive all interference and minimize their transmission of interference. All of these rules are about interference avoidance and mitigation. They also protect licensed users above all others...like if you point a relatively small transmitter at a satellite's transponder and jam it you might go to jail.

In reality this is all ridiculous. Modern radio technologies are able to use any part of the spectrum as efficiently as possible and "tune around" interference. Ultra Wideband radios can pull transmissions out from below the noise floor on many bands, so that most users can't even hear them because they don't know what to look for. This stuff is here now and has made licensing a tool for incumbent monopolies and not really a technical issue.

So we're at a point in our advancement where the technology has once again surpassed the usefulness of the system that grew up to regulate it. Too bad there are billions (trillions?) of dollars tied up in this old system that is essentially meaningless.


Reading through
http://hraunfoss.fcc.gov/edocs_public/attachmatch/FCC-10-174A1.pdf and
if anyone else knows what is possible now for packet forwarding, using
radio relays in unlicensed spectrum etc please share here. Thanks!

In general you can use repeaters in ISM or White Spaces bands provided you adhere to power and antenna height requirements.

As I said above this is technically moot, but we play the game anyway.

--
Aaron Huslage -  +1-919-600-1712
http://blog.hact.net
IM: AIM - ahuslage; Yahoo - ahuslage; MSN - hus...@gmail.com; GTalk - hus...@gmail.com

Colin Hawkett

unread,
Feb 25, 2011, 8:39:35 PM2/25/11
to building-a-distributed...@googlegroups.com, Colin Hawkett

Definitely, to some degree, I'm trying to clarify the boundaries. More comments below.


On Friday, February 25, 2011 7:11:20 PM UTC, Sam Rose wrote:
For myself, personally:

On Fri, Feb 25, 2011 at 1:47 PM, Colin Hawkett <haw...@gmail.com> wrote:
> This might not be the most popular perspective, but I'm not sure everybody
> is trying to solve the same problem here, or for the same reasons. There's
> some complicated socio-political agendas and different people see different
> stuff - the path from A->B is assumed obvious - indeed that A and B are
> obvious. It might just be me who's not sure :)
> Is it that we want to be able to use the web we know today on a
> decentralised hardware and software infrastructure? Why?

Yes, this is *part* of it. We should be able to take existing open
protocols and web technologies and run them on a distributed internet.
Why? It doesn't make sense to me to try and leave that behind right
now, plus it can work with what will very likely be a reasnable
investment on targetted development in the traditional web application
"stack". Some emerging web applications may work right out of the box
(couchdb etc etc)

I'm not so sure about this. Again, it depends what we want to achieve. My perspective is that the very best distributed stuff we do right now is in data centres, where bandwidth between nodes is massive, and we can handle the coherence stuff on high volatility data. Those guys really can plug anther node in and get N + 1 benefit. If it was possible, we would do it between datacentres, and then between individual computers, but it is just not feasible with our current technology. I have no doubt Google is gearing up Chrome (the browser) to be the software on a distributed node, which is why they spend so much energy evangelising improvements to the physical layers.


There is a subset of problems for which we can achieve fine grained distributed (lower volatility data gets easier, as does data for which very eventual consistency is ok), but that really is the question - is that subset sufficient? There's some discussion about techies and non-techies here, so I think we need to be clear that the ideal of the distributed internet from a philosophical perspective and the reality of the technology at this point in time are different things. We also need to be clear that a distributed internet won't behave the same as our current internet. e.g. latency, reliability and performance profiles will noticeably be different. 

> Which parts of that
> problem are we proposing to solve? By what path? To what extent? For what
> benefit?

In the context of contactcon, we want to keep the floor open, we want
to support the various problems people are trying to solve in
parallel. I don't want to just focus in on one problem, because that
will exclude others that I think are important. There are a plurality
of problems. That might not seem helpful for focusing in on "what can
be done now specifically" but let's bring some more players to the
discussion, and more people working on more parts of the picture
before we try to narrow down the focus, is what I personally think.
We're in the process of talking directly with people working on the
hardware, protocol and software layers now.

A plurality of problems is ok, as long as we define what they are.

I think the next stage is to get those folks here to answer all of
these questions we're asking. Then, when there more stakeholders, we
can draw some boundaries which will narrow things down. Just my
opinion. I think we need at least another week or 2 before we start
focusing in/narrowing down, if only to give us a chance to get more
stakeholders in the discussion. I am open minded though, and this is
only my personal take on it.

> Decentralised is an ambiguous terminology. e.g. is a local community
> wireless transmitter centralised or decentralised?
> Basically what's the problem, what are the options, what are their
> advantages/disadvantages, on what time scales will they manifest, under what
> constraints & assumptions. What are the outcomes we are looking for, and can
> we measure success/failure?
> A distributed internet isn't an outcome - it facilitates an outcome. What is
> that?
>

I know many people who probably would disagree that a distributed
internet is not an outcome. Although they might agree that it
facilitates many outcomes. I think we've documented thousands of those
outcomes that a distributed internet facilitates at
http://p2pfoundation.net throughout that wiki over the last 5 years.

Take the point re. distributed internet not being an outcome :) It reminds me though of the monty python sketch where Arthur offers to let the French join him in their quest for the grail. And they say they've already got one. Depending on perspective, some might say we've already got one. Are they wrong? It depends entirely on perception. Labouring the analogy a little further, the quest for the grail delivers the rewards, not the grail itself :)

Still, I understand your frustration, and reasoning for wanting to
focus in. Personally, I'd like to see a distributed hardware/network
medium where people can pool hardware resources and network connection
node resources to run some variant of practically anything they can do
on the internet now (starting with the simplest things, like message
passing, etc).

It's good to see a technical use-case :)

Cheers,

Colin

Robert Steele

unread,
Feb 25, 2011, 8:45:58 PM2/25/11
to building-a-distributed...@googlegroups.com
Colin, your message really resonates with me, and I see a huge need to get all this back and forth into the P2P wiki.

I have been thinking about the following, and my wake up call from Doug and Venessa was to realize 1 comes before 2 (what I've been working on for 20 years).

1.  Autonomous Internet access for the three billion poor as well as all of us (hand-held emphasis)

2.  Global virtual networks of ad hoc educators, translators, and support people with desktop access to Internet (Desktop emphasis)

3.  Regional multinational intelligence centers and call centers that aggregate all incoming, educate the poor one cell call at a time, and do massive near real time processing that can be drawn on by anyone--including massive SMS data reception and self-making.

Robert

Colin Hawkett

unread,
Feb 25, 2011, 9:14:10 PM2/25/11
to building-a-distributed...@googlegroups.com
I'm thinking we should be able to get most of this with our current infrastructure, no?

Eugen Leitl

unread,
Feb 26, 2011, 5:07:07 AM2/26/11
to building-a-distributed...@googlegroups.com
On Fri, Feb 25, 2011 at 03:02:50PM -0800, Venessa Miemis wrote:
> hey eugen!
>
> sorry if i haven't put a place on this google group yet that explains
> the intentions..... i'm organizing a gathering in nyc in october with
> 400 of the top people i can find in the world who are working on the
> various aspects of the emerging networked society. i started this
> group to get some conversation and action going. one of the
> initiatives that is coming together here is to talk about distributed
> internet infrastructures. because we are promoting open discussion and
> collaborative solutions, this is a place where the big picture takes
> precedent over distinctions over theory verse practice. i think this

You want a magic box where TCP/IP comes out. The box and juice
comes out of your own pocket. The box is zero administration,
and talks to other boxes all over the world. Entire traffic
is in cypher, so no way to sniff or block. The default mode
no egress point to the Internet, nor easy ability to localize
the service published by the box. The network must be resilient
to abuse.

This is the level of theory that suggests practice (implementation).
Most of above problems have been solved, some not. Some
problems are trivial, some hard.

> movement as a whole has been stunted for years because it's so

Try decades. At least three.

> splintered. every faction seems to want to claim that its aspect is

I would say there never was a faction with a budget. You
do not need a lot of money, particularly today, but you
do need some money.

> the essential one. i think it's worth a try to be inclusionary and

A workable standard is typically not made by a commitee.

> understand that a problem of this scale and complexity is going to
> need A LOT of intelligent brains working together. i think it's gonna

You need less than 10 engineers, but they need to be top people.
The only reason this is feasible is because many facets of the
problem have been already solved.

Adding more and lesser people will guarantee that you will not
succeed.

> be tough at first to figure out how to find some common language
> between the builders and the thinkers and all the various people
> involved, but i think we can.
>
> please contribute your voice and your opinions and your community and
> your work... but please don't prevent others to bring their ideas,
> concerns, and projects to the table as well.

Without ability to muffle voices you will fail.



> i think we have a chance here to see how all of our skills and
> knowledge can fit together in a bigger way.

--

Eugen* Leitl <a href="http://leitl.org">leitl</a> http://leitl.org
______________________________________________________________

ICBM: 48.07100, 11.36820 http://www.ativel.com http://postbiota.org

Robert Steele

unread,
Feb 26, 2011, 9:35:49 AM2/26/11
to building-a-distributed...@googlegroups.com
Leitl for President.  This and your other posts are  "core" (for me) and I would love to see you be the one that puts all this up
as a starter of the Wiki Page that I have suggested be named Open Source Everything Road-Map.  We need to get this
wonderful dialog into the wiki.

Here is my own concept, for a Global Range of Gifts table that allows the one billion rich to fund the connection of the
three billion poor.  By name giving, micro giving where Autonomous Internet facilitation has more karma points.  We
do not lack for money, we lack of the orchestration of tiny amounts of money.

I had the chance to work with Joe Trippi and we designed a $500M public intelligence network in which each of 100
million Americans was asked to give just five dollars a year as a subscription, so as to match the spending of the
corporations and fight with our own money in the aggregate against the corruption and diminuition of the US electoral
process, which is so dirty now we don't equal qualify as a full democracy by international standards.

Now imagine a billion people giving an average of five dollars each.

2012 is the year of awakening.  I think this group and what it can produce in the next 90 days is central to the future.
IMHO.

Venessa Miemis

unread,
Feb 26, 2011, 10:48:34 AM2/26/11
to Building a Distributed Decentralized Internet
summary of above thread to be transferred to wiki:

Q1. Is it that we want to be able to use the web we know today on a
decentralized hardware and software infrastructure? Why?

A1. Yes, we should be able to take existing open protocols and web
technologies and run them on a distributed internet. It can work with
what will very likely be a reasonable
investment on targetted development in the traditional web application
"stack". I'd like to see a distributed hardware/network medium where
people can pool hardware resources and network connection node
resources to run some variant of practically anything they can do on
the internet now (starting with the simplest things, like message
passing, etc).

A1:Counterpoint. The very best distributed stuff we do right now is
in data centres, where bandwidth between nodes is massive, and we can
handle the coherence stuff on high volatility data. Those guys really
can plug anther node in and get N + 1 benefit. If it was possible, we
would do it between datacentres, and then between individual
computers, but it is just not feasible with our current technology.

There is a subset of problems for which we can achieve fine grained
distributed (lower volatility data gets easier, as does data for which
very eventual consistency is ok), but that really is the question - is
that subset sufficient? There's some discussion about techies and non-
techies here, so I think we need to be clear that the ideal of the
distributed internet from a philosophical perspective and the reality
of the technology at this point in time are different things. We also
need to be clear that a distributed internet won't behave the same as
our current internet. e.g. latency, reliability and performance
profiles will noticeably be different.

Q2. What are existing US rules (and EU rules too) around radio
A2. EU has no white spaces yet, as far as I'm aware. The US recently
announced its rules so white spaces protocol finalization and use
should begin in the next year or two.
FCC rules:
There are Part-15 bands (also called ISM) where any unlicensed user
can transmit up to a ceiling of power (100mW for 2.4gHz, etc). These
devices must receive all interference and minimize their transmission
of interference. All of these rules are about interference avoidance
and mitigation. They also protect licensed users above all
others...like if you point a relatively small transmitter at a
satellite's transponder and jam it you might go to jail.

Modern radio technologies are able to use any part of the spectrum as
efficiently as possible and "tune around" interference. Ultra Wideband
radios can pull transmissions out from below the noise floor on many
bands, so that most users can't even hear them because they don't know
what to look for. This stuff is here now and has made licensing a tool
for incumbent monopolies and not really a technical issue.

Q3. What is possible now for packet forwarding, using radio relays in
unlicensed spectrum? (Reading through
http://hraunfoss.fcc.gov/edocs_public/attachmatch/FCC-10-174A1.pdf)

A3. In general you can use repeaters in ISM or White Spaces bands
provided you adhere to power and antenna height requirements. This is
technically moot.

Joaquín Salvachúa

unread,
Feb 26, 2011, 12:04:39 PM2/26/11
to building-a-distributed...@googlegroups.com

El 26/02/2011, a las 16:48, Venessa Miemis escribió:

> summary of above thread to be transferred to wiki:
>
> Q1. Is it that we want to be able to use the web we know today on a
> decentralized hardware and software infrastructure? Why?
>
> A1. Yes, we should be able to take existing open protocols and web
> technologies and run them on a distributed internet. It can work with
> what will very likely be a reasonable
> investment on targetted development in the traditional web application
> "stack". I'd like to see a distributed hardware/network medium where
> people can pool hardware resources and network connection node
> resources to run some variant of practically anything they can do on
> the internet now (starting with the simplest things, like message
> passing, etc).
>
> A1:Counterpoint. The very best distributed stuff we do right now is
> in data centres, where bandwidth between nodes is massive, and we can
> handle the coherence stuff on high volatility data. Those guys really
> can plug anther node in and get N + 1 benefit. If it was possible, we
> would do it between datacentres, and then between individual
> computers, but it is just not feasible with our current technology.
>

Actual web is based on HTTP protocol. This may run over any TCP or point to point transport protocol. Is perfectly isolated and any non IP addressing schema may be used URL-URI http/html links. HTML (even html5) is fully
independant from the transmission protocol. Just supposes the underlaying protocol is fully reliable, that is no failure semantic into HTML.

You must be aware that change IP is really tough to made something better. Also on routing procotols. All yours ideas are about border routing, nothing about how core routing (or how it may be funded).

We in EU is quite hard to transmit information without having a licence. For an ex-ISO implementer european experience this is tough.


> Q3. What is possible now for packet forwarding, using radio relays in
> unlicensed spectrum? (Reading through
> http://hraunfoss.fcc.gov/edocs_public/attachmatch/FCC-10-174A1.pdf)
>
> A3. In general you can use repeaters in ISM or White Spaces bands
> provided you adhere to power and antenna height requirements. This is
> technically moot.
>

Joaquin Salvachua

Robert Steele

unread,
Feb 26, 2011, 12:10:23 PM2/26/11
to building-a-distributed...@googlegroups.com
Core Routing---perfect sub-page.  Working on creating the pages, not the content.  Back to everyone in a couple of hours.

Eugen Leitl

unread,
Feb 26, 2011, 2:01:17 PM2/26/11
to Samuel Rose, building-a-distributed...@googlegroups.com
On Fri, Feb 25, 2011 at 06:11:45PM -0500, Samuel Rose wrote:

> Paul B. Hartzog, Richard C. Adler and myself have written an RFC.
> We're not "soft humanities people".

The only Rose on http://www.rfc-editor.org/cgi-bin/rfcsearch.pl
is Marshall. The only Adler is Adler-32 checksum
http://www.rfc-editor.org/rfc/rfc3309.txt

There is no Hartzog. The only thing vaguely resembling an
RFC I can find is draft for something like Flows. Is that you?

http://flows.panarchy.com/index.php?title=FlowsRFCDraft



> I don't think we are at the stage in *this* discussion to start
> writing RFC's, personally. Although I do agree that writing RFC(s)
> could be an early activity that we could take on doing very soon.

I think ability to read RFCs and technical design documents in
general is a skill many participants here need to be pick up.

I agree that the ability to write RFCs is currently unnecessary, but
it will become so assuming this project will ever succeed.

Eugen Leitl

unread,
Feb 26, 2011, 2:13:32 PM2/26/11
to building-a-distributed...@googlegroups.com
On Fri, Feb 25, 2011 at 06:51:13PM -0500, Devin Balkind wrote:
> Eugen. Enthusiastic students do amazing work so if you'd like to talk
> technical, please post your documentation and we will start following
> your rss at floing and provide you support.

Do you have access to indentured labor, aka C.S. undergrads
and graduates? Ability to run large scale network simulations
is crucial in order to find design flaws and measure
scaling. There's plenty to build and to publish, so academic
ties would be good.

> If someone doesn't start a Tiddlywiki for this group before tomorrow I
> will start one at tiddlyspot.
>
> The other things people have posted on this board have been very
> exciting and look forward to posting more when I get off the road
> tomorrow.
>
> DWB

--

Eugen* Leitl <a href="http://leitl.org">leitl</a> http://leitl.org
______________________________________________________________

ICBM: 48.07100, 11.36820 http://www.ativel.com http://postbiota.org

Eugen Leitl

unread,
Feb 26, 2011, 2:19:35 PM2/26/11
to building-a-distributed...@googlegroups.com
On Fri, Feb 25, 2011 at 06:57:31PM -0500, Robert Steele wrote:

> I see satellites, and nano-satellites,

If it's LEO you need a large constellation, or settle
for a few birds with DTN onboard, and according ground
infrastructure which knows about ephemerides and can
track birds either by parabol dish or by phased array
antenna. In any case this is well over 10 MUSD budget.

> and WiMax, and pigions, all coming

Rather WiFi, and 3G and 4G.

> into play.
> "Whatever it takes" to include breaking spectrum rules that are antiquated,

It is very easy to produce a tragedy of the commons with a limited
resource like spectrum. Play nice.

> when
> necessary for in extremis support to good causes.
>
> The US Government is so hosed on spectrum they cannot run half the stuff
> they
> have in Afghanistan without spectrum conflicts. This is a good time for a
> breakout, IMHO.

The best way is to use extremely directional line of sight links, and
higher frequencies.

--
Eugen* Leitl <a href="http://leitl.org">leitl</a> http://leitl.org
______________________________________________________________

ICBM: 48.07100, 11.36820 http://www.ativel.com http://postbiota.org

Paul B. Hartzog

unread,
Feb 26, 2011, 3:14:57 PM2/26/11
to building-a-distributed...@googlegroups.com, Colin Hawkett
What an interesting thread. Some useful, some way off in left field,
some cooperative, some perverse Orwellian, and more.

Colin,
Thanks for trying to keep things focussed.
I've been on vacation but I'll try to catch up and respond. Here goes.

On Fri, Feb 25, 2011 at 1:47 PM, Colin Hawkett <haw...@gmail.com> wrote:
> This might not be the most popular perspective, but I'm not sure everybody
> is trying to solve the same problem here, or for the same reasons. There's
> some complicated socio-political agendas and different people see different
> stuff - the path from A->B is assumed obvious - indeed that A and B are
> obvious. It might just be me who's not sure :)

Much that seems obvious to some will need to be articulated to others.
These articulations will need to be in different ways (language),
and at different times (repetition),
because people learn differently.

> Is it that we want to be able to use the web we know today on a

> decentralised hardware and software infrastructure? Why? Which parts of that


> problem are we proposing to solve? By what path? To what extent? For what
> benefit?

> Decentralised is an ambiguous terminology. e.g. is a local community
> wireless transmitter centralised or decentralised?

Yes, language is crucial.
An old metaphor I have used in the past is the ATM, i.e.
does it represent centralisation or decentralisation?
The point is that old terms OFTEN no longer apply
when we enter a space of new meanings
like we are currently doing.
New language will emerge. New discourse.

> Basically what's the problem, what are the options, what are their
> advantages/disadvantages, on what time scales will they manifest, under what
> constraints & assumptions. What are the outcomes we are looking for, and can
> we measure success/failure?
> A distributed internet isn't an outcome - it facilitates an outcome. What is
> that?

Infrastructure can be thought of as an outcome of course,
but it also facilitates other outcomes, and works against others.
However, it is also true that infrastructure is not an outcome,
no more than a species is an 'outcome' of evolution.
Infrastructure that perpetually evolves is the goal here.

Perpetual evolution means
1) don't do anything to produce or 'lock-in' vis a vis standards
Lock-in is like over-adaptation that becomes maladaptation.
2) vigilantly protect the openness necessary for new innovations
New innovations are the only way to keep from getting stuck
in overly-adapted maladaptive optima in which everything
is locked in to everything else.

hope that helps

-p

--------------------------------------------------------
The Forward Foundation
http://www.ForwardFound.org

paul.b....@forwardfound.org
--------------------------------------------------------
http://www.PaulBHartzog.org
PaulBH...@PaulBHartzog.org
--------------------------------------------------------
http://www.panarchy.com
PaulBH...@panarchy.com
--------------------------------------------------------
University of Michigan
PHar...@umich.edu

Eugen Leitl

unread,
Feb 26, 2011, 3:18:27 PM2/26/11
to building-a-distributed...@googlegroups.com
On Fri, Feb 25, 2011 at 06:14:10PM -0800, Colin Hawkett wrote:
> I'm thinking we should be able to get most of this with our current
> infrastructure, no?

Yes, the idea is minimal disruption, and and slow growth,
providing added value right from the start.

Much would be helped with just bundling a number of
existing packages into an appliance. It will have to
be able to autoupdate from own cloud (both firmware
image and packages) by default.

Aaron Huslage

unread,
Feb 26, 2011, 3:25:39 PM2/26/11
to building-a-distributed...@googlegroups.com

A lot of eugene's ideas seem to be encompassed in the freedombox project. I have issues with this approach, but I'm ok with being wrong.

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

Eugen Leitl

unread,
Feb 26, 2011, 3:33:39 PM2/26/11
to building-a-distributed...@googlegroups.com
On Sat, Feb 26, 2011 at 03:25:39PM -0500, Aaron Huslage wrote:

> A lot of eugene's ideas seem to be encompassed in the freedombox project. I

It's for a reason, the people around Eben Moglen are not stupid.
There are not too many ways in how you can make the project
a success.

> have issues with this approach, but I'm ok with being wrong.

What are your issues?

Colin Hawkett

unread,
Feb 26, 2011, 5:36:54 PM2/26/11
to building-a-distributed...@googlegroups.com, Colin Hawkett

Much that seems obvious to some will need to be articulated to others.
These articulations will need to be in different ways (language),
and at different times (repetition),
because people learn differently.

I'm worried there is an assumption of universal truth here, that those who perceive it must communicate it to the uninformed? I guess I'm really just looking for 

Problems: X, Y, Z
Solution: P2P something or other

so that we can discuss whether we're even chasing the right end game.

Yes, language is crucial.
An old metaphor I have used in the past is the ATM, i.e.
does it represent centralisation or decentralisation?
The point is that old terms OFTEN no longer apply
when we enter a space of new meanings
like we are currently doing.
New language will emerge.  New discourse.

I'm not sure the lack of definition is a result of the lack of language. I'm just trying to avoid the kool-aid.

Infrastructure can be thought of as an outcome of course,
but it also facilitates other outcomes, and works against others.
However, it is also true that infrastructure is not an outcome,
no more than a species is an 'outcome' of evolution.
Infrastructure that perpetually evolves is the goal here.

Perpetual evolution means
1) don't do anything to produce or 'lock-in' vis a vis standards
Lock-in is like over-adaptation that becomes maladaptation.
2) vigilantly protect the openness necessary for new innovations
New innovations are the only way to keep from getting stuck
in overly-adapted maladaptive optima in which everything
is locked in to everything else.

I think this is an impossible goal. The most amazing tech we have today (the internet) is a triumph of standards over proprietary fragmentation. Ok I'm gonna get all philosophical here, so fair warning.

There is a balance (as there is in all things) between the extremes of diversity and ubiquity. To say 'no lock-in' focuses solely on diversity. All diversity requires a strata of ubiquity on which to thrive. And indeed, that strata must be constantly enriched by the elements of supported diversity which become dominant through natural selection. For example, on the web, we have reached the point where the social graph (among other things) needs to become part of the ubiquitous infrastructure on which yet even greater diversity can thrive. Diversity does not exist in isolation. We need to work in harmony with that process.

Paul B. Hartzog

unread,
Feb 26, 2011, 6:47:17 PM2/26/11
to building-a-distributed...@googlegroups.com, Colin Hawkett
I think we're arguing the same thing here:

"There is a balance (as there is in all things) between the extremes
of diversity and ubiquity."

The "edge of chaos" balance that many complex systems are capable of achieving,
is possible only with openness and continuing diversity.

In other words "too little diversity" will kill a population,
but
"too much diversity" simply self-corrects back to sustainable equilibrium.

The math is the same as the carrying-capacity graph
with upper and lower thresholds, here:
http://blog.p2pfoundation.net/understanding-carrying-capacity/2010/07/01

Definitely want the balance though.

-p

--

Samuel Rose

unread,
Feb 26, 2011, 6:50:53 PM2/26/11
to Eugen Leitl, building-a-distributed...@googlegroups.com
On Sat, Feb 26, 2011 at 2:01 PM, Eugen Leitl <eu...@leitl.org> wrote:
> On Fri, Feb 25, 2011 at 06:11:45PM -0500, Samuel Rose wrote:
>
>> Paul B. Hartzog, Richard C. Adler and myself have written an RFC.
>> We're not "soft humanities people".
>
> The only Rose on http://www.rfc-editor.org/cgi-bin/rfcsearch.pl
> is Marshall. The only Adler is Adler-32 checksum
> http://www.rfc-editor.org/rfc/rfc3309.txt
>
> There is no Hartzog. The only thing vaguely resembling an
> RFC I can find is draft for something like Flows. Is that you?
>
> http://flows.panarchy.com/index.php?title=FlowsRFCDraft
>

Yes, this is the RFC I am referring to.

>> I don't think we are at the stage in *this* discussion to start
>> writing RFC's, personally. Although I do agree that writing RFC(s)
>> could be an early activity that we could take on doing very soon.
>
> I think ability to read RFCs and technical design documents in
> general is a skill many participants here need to be pick up.
>
> I agree that the ability to write RFCs is currently unnecessary, but
> it will become so assuming this project will ever succeed.
>


Yes agree 100% Let's say that this definitely one of the first things
we can do to move concretely from discussion to action. We may end up
creating a new email list, or possibly doing it here, but
reading/writing RFC's for participant/proposed technologies is
something we can and should support in the coming weeks and months.
Hopefully we can start this activity in around 2 weeks or so.

Eugen, we can really use your help with this when the time comes. If
you happen to identify where there is an immediate need to identify
existing RFC or write new ones, please don't hesitate.

> --
> 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
>

--

Samuel Rose

unread,
Feb 26, 2011, 7:18:04 PM2/26/11
to building-a-distributed...@googlegroups.com, Venessa Miemis
On Sat, Feb 26, 2011 at 10:48 AM, Venessa Miemis
<veness...@gmail.com> wrote:
> summary of above thread to be transferred to wiki:
>
http://p2pfoundation.net/Category:P2P_Infrastructure#Q_.26_A_from_Building_a_Distributed_and_Decentralized_Internet

added to URL above for now, until we figure out a way to refactor this all.

Robert Steele

unread,
Feb 26, 2011, 7:30:22 PM2/26/11
to building-a-distributed...@googlegroups.com, Samuel Rose, Venessa Miemis
please note that the P2P infrastructure is embedded in the table of contents for the Open Everything Road-Map page.

my sense is that there is a need for a separate Autonomous  Internet Assembly Guide that sticks to the specifics.  what I see at P2P infrastructure is all the guides, discussions, etcetera.  would be nice to have a "manual" that is nice and clean and really focused.

Colin Hawkett

unread,
Feb 26, 2011, 7:38:23 PM2/26/11
to building-a-distributed...@googlegroups.com, Colin Hawkett
I think you're probably right - although I would be wary of an approach avoiding the standard platform - we should be engineering it, rather than fighting against its emergence. I always end up back at the internet as an example - if there were diverse internets, then we wouldn't have the connective capabilities we have today. In some ways capitalist competition theory promotes diversity, and I'm not sure it is always a good thing. I'm not arguing monopoly is good, more that if we can engineer the standard platform well, and associated governance, the diversity that springs from it is far greater than if we had fought to keep the platform itself diverse. Ubiquity begets diversity, begets ubiquity.

>> paul....@forwardfound.org
>> --------------------------------------------------------
>> http://www.PaulBHartzog.org
>> Paul...@PaulBHartzog.org
>> --------------------------------------------------------
>> http://www.panarchy.com
>> Paul...@panarchy.com
>> --------------------------------------------------------
>> University of Michigan
>> PHa...@umich.edu

Robert Steele

unread,
Feb 26, 2011, 7:45:39 PM2/26/11
to building-a-distributed...@googlegroups.com
Clarity in sharing of true cost information is an enabler that leverages diversity while avoiding discordance.

If Michel's site is robust enough to welcome all comers, disparate efforts can be informally harmonized through non-intrusive information sharing.

Samuel Rose

unread,
Feb 26, 2011, 7:47:33 PM2/26/11
to Robert Steele, building-a-distributed...@googlegroups.com, Venessa Miemis
On Sat, Feb 26, 2011 at 7:30 PM, Robert Steele
<robert.david...@gmail.com> wrote:
> please note that the P2P infrastructure is embedded in the table of contents
> for the Open Everything Road-Map page.
>
> my sense is that there is a need for a separate Autonomous  Internet
> Assembly Guide that sticks to the specifics.  what I see at P2P
> infrastructure is all the guides, discussions, etcetera.  would be nice to
> have a "manual" that is nice and clean and really focused.
>


Robert,

If you can either email back to me here, or put on P2PF wiki somewhere
the collection of pages you've been suggesting in these threads, I'll
work on those over the next few days, and try to come up with a
workflow for building them over time.

I think so far we have:

http://p2pfoundation.net/Category:P2P <- this one seems to be a good
place to keep really clean as a page for everyone to look at and
understand the space, technical people or not

http://p2pfoundation.net/Category:P2P_Infrastructure <- I am using
this one now as a drafting and parking space. I will ultimately clean
it back up into what it was before, but will for now park content
there.

Are there other pages we should be aware of?

Also I like the idea of

Autonomous  Internet Assembly Guide <- we can start this as a kind of
practical publication-formatted page, and we can make sure that Devin
and those who want to transfer that to Tiddlywiki spaces will be able
to do so as easily as can be done.

Robert Steele

unread,
Feb 26, 2011, 7:56:22 PM2/26/11
to Samuel Rose, building-a-distributed...@googlegroups.com, Venessa Miemis
http://p2pfoundation.net/Category:P2P <- this one seems to be a good
place to keep really clean as a page for everyone to look at and
understand the space, technical people or not

Sam, this is the page I created today with content, after Michel created it as a blank page.
This is the page that I am hoping will be easy to get to from the home page, have asked
Michel to indent Open Source Everything Road-Map under Open Source Everything

For the convenience of others, the table of contents (no content yet) is at the end of this message


http://p2pfoundation.net/Category:P2P_Infrastructure <- I am using
this one now as a drafting and parking space. I will ultimately clean
it back up into what it was before, but will for now park content
there.

This is the original page, I was careful to include it as the first page under technical but
I see so much thrown in here it just does not seem to be the right place for the
component level specifics.

NOT YET CREATED is a page that could be called Guide to Building an Autonomous Internet,
that could be the clean manifestation of concensus, in other words, the technical pages contain
everything whether the group agrees on not, while the Guide is a consensus offering of
"best available interoperable open stuff."

Contents

[hide]


Samuel Rose

unread,
Feb 26, 2011, 8:15:35 PM2/26/11
to Robert Steele, building-a-distributed...@googlegroups.com, Venessa Miemis
On Sat, Feb 26, 2011 at 7:56 PM, Robert Steele
<robert.david...@gmail.com> wrote:
> http://p2pfoundation.net/Category:P2P <- this one seems to be a good
> place to keep really clean as a page for everyone to look at and
> understand the space, technical people or not
>
> Sam, this is the page I created today with content, after Michel created it
> as a blank page.
> This is the page that I am hoping will be easy to get to from the home page,
> have asked
> Michel to indent Open Source Everything Road-Map under Open Source
> Everything
>


Ok, great! So, this may seem kind of dense, but bear with me:

How do you imagine we will populate this page you created today. What
is the process you imagine? I am willing to work with you on it, just
want to get on the same page with you on what you imagine here.

> For the convenience of others, the table of contents (no content yet) is at
> the end of this message
>

I think so far it is a good table of contents!

> http://p2pfoundation.net/Category:P2P_Infrastructure <- I am using
> this one now as a drafting and parking space. I will ultimately clean
> it back up into what it was before, but will for now park content
> there.
>
> This is the original page, I was careful to include it as the first page
> under technical but
> I see so much thrown in here it just does not seem to be the right place for
> the
> component level specifics.
>

Yes, again, I have been using it as a place to park things that are
destined for elsewhere, and will eventually clean it back down to what
it was. The real question is: how to refactor that stuff into pages we
can use? What is our process?

> NOT YET CREATED is a page that could be called Guide to Building an
> Autonomous Internet,
> that could be the clean manifestation of concensus, in other words, the
> technical pages contain
> everything

What are the "technical pages". The headings/table of contents on
http://p2pfoundation.net/Category:P2P ? If so, what is a simple way
for us to begin adding to those pages from the discussion taking place
here?

> whether the group agrees on not, while the Guide is a consensus
> offering of
> "best available interoperable open stuff."
>

Ok, this all sounds great to me. I just want to work with you to
create a process that allows us to build from discussion here and
elsewhere into these pages. Maybe we'll need some intermediary pages,
or maybe we'll need a criteria of some sort (such as having references
for assertions, finding and referencing existing RFC docs, etc etc)
Let's talk about this and maybe come up with a process by monday, ok?

> Contents
>
> [hide]
>
> 1 Strategic Overview
>
> 1.1 Connecting Humanity
> 1.2 Technical Terms of Reference
> 1.3 Financial Objectives
>
> 2 Operational Overview
>
> 2.1 Empowering the Billions of Poor
> 2.2 Open Source Foundation Technical Map
> 2.3 Financial Possibilities
>
> 3 Tactical Overview
>
> 3.1 Pilot Projects
> 3.2 Important Gaps in Open Capability
> 3.3 Financial Needs
>
> 4 Technical Overview
>
> 4.1 P2P Infrastructure
> 4.2 Transport Layer
> 4.3 Router Layer
> 4.4 Host Layer
> 4.5 Application Layer
> 4.6 Matrix of Financial Costs
>
>

--

Robert Steele

unread,
Feb 26, 2011, 8:37:03 PM2/26/11
to Samuel Rose, building-a-distributed...@googlegroups.com, Venessa Miemis
This may sound simplistic, but I think we have reached a point of open contribution.

I think I get out of the way now, and each person migrates their own stuff to the wiki as they see fit.

As a general rule, and this is the humanist editor talking, not the technical manual writer:

top level page: limit insertions to one short paragraph with its own sub-heading, use that to point to new internal page or external link as appropriate.

create all the sub pages and sub sub pages and uploaded files desired so as to build out to where there is a sub page for every individual software, every individual piece of hardware, then intermediate pages for how they fit together, and finally a short blurb with pointers on the top level pages.

the P2P infrastructure page seems cluttered, but that is not my area of  expertise, so i just throw out the comment that by using the top level pages, especially Transport, Router, and Host, there is an opportunity to break stuff out into its own pages and then provide short connecting blurbs

what I have tried to provide is a top level structure that distinguishes between the four levels of strategy, operations, tactics, and technical, while providing for each level a place for human factors, technical, and financial.

beyond that, I think it is up to the wizards to populate and then we "artists" come in later with a dust mop.

does that make sense?  I have a wonderful feeling of expectation about this, I think Doug and Venessa have catalyzed a very fine gathering at just the right time, Egypt and the Internet work-arounds discussion that Egpt sparked are in my view historic.

I have no "pride of authorship" on the top level tables, deferring to all of you as to whether that structure serves the group.  I do have in the back of my mind business plans and funding proposals I have written, the organization is intended to cover the big three:  people, things, and money.

Connexus Sumus Unum.....if anyone knows a Latin wizard, would be glad to have this checked, intent is "Connected, We Are One"

Aaron Huslage

unread,
Feb 26, 2011, 8:45:48 PM2/26/11
to building-a-distributed...@googlegroups.com, Eugen Leitl
Rewinding this discussion to a time when it seemed useful...

On Sat, Feb 26, 2011 at 3:33 PM, Eugen Leitl <eu...@leitl.org> wrote:

> have issues with this approach, but I'm ok with being wrong.

What are your issues?
 
From what I can tell, they could simply co-opt much of the work that was done on OLPC's server and client and spit out a linux distro. Not to mention the DD-WRT work. We have this stuff pre-baked and throwing money at the same solution over and over when it just doesn't work is not all that useful to me.

Wireless Mesh TCP/IP networks simply are not usable, scalable or connectable enough to work. We've been trying this for close to 10 years on the Wi-Fi incarnation and it's just beating our heads against 20 walls over and over again. They ALL require some upstream carrier to be a backhaul to the internet, none of them distribute outbound routes properly, they all get heavily congested at a certain point with routing traffic and metadata.

Not to mention that wireless mesh networks are incredibly easy to find and are totally inappropriate for sending into surveillance states like Libya or Iran...as I've seen get funded lately. The name Wireless Mesh on a grant application is damn near guaranteed to make you attractive. This is dangerous and stupid and people will get killed. That makes me sad.

I'm not saying I have an answer here. I just know what the solution isn't. These networks aren't p2p and they don't work.

--
Aaron Huslage -  +1-919-600-1712
http://blog.hact.net
IM: AIM - ahuslage; Yahoo - ahuslage; MSN - hus...@gmail.com; GTalk - hus...@gmail.com

Samuel Rose

unread,
Feb 26, 2011, 8:46:19 PM2/26/11
to Robert Steele, building-a-distributed...@googlegroups.com, Venessa Miemis
On Sat, Feb 26, 2011 at 8:37 PM, Robert Steele
<robert.david...@gmail.com> wrote:
> This may sound simplistic, but I think we have reached a point of open
> contribution.
>
> I think I get out of the way now, and each person migrates their own stuff
> to the wiki as they see fit.
>
> As a general rule, and this is the humanist editor talking, not the
> technical manual writer:
>
> top level page: limit insertions to one short paragraph with its own
> sub-heading, use that to point to new internal page or external link as
> appropriate.
>
> create all the sub pages and sub sub pages and uploaded files desired so as
> to build out to where there is a sub page for every individual software,
> every individual piece of hardware, then intermediate pages for how they fit
> together, and finally a short blurb with pointers on the top level pages.
>
> the P2P infrastructure page seems cluttered, but that is not my area of
> expertise, so i just throw out the comment that by using the top level
> pages, especially Transport, Router, and Host, there is an opportunity to
> break stuff out into its own pages and then provide short connecting blurbs
>

This actually makes a lot of sense. So
http://p2pfoundation.net/Category:P2P really just becomes pointers to
other pages. Ok, works for me. We may need to *create* some of these
pages, but yeah, I think this will work.

> what I have tried to provide is a top level structure that distinguishes
> between the four levels of strategy, operations, tactics, and technical,
> while providing for each level a place for human factors, technical, and
> financial.
>
> beyond that, I think it is up to the wizards to populate and then we
> "artists" come in later with a dust mop.
>

Well, I've got lots to write in the wiki on a technical level for
sure. But, this gives us a way to get started, that I think can work
well. I'll begin refactoring my work and other contribs from P2P
Infrastructure page to new pages that are linked to from
http://p2pfoundation.net/Category:P2P this in the coming week.

I think supporting citations and related existing RFC's (if any) could
be useful on all of these subpages, too. I may make a template of some
type for that eventually.

> does that make sense?  I have a wonderful feeling of expectation about this,
> I think Doug and Venessa have catalyzed a very fine gathering at just the
> right time, Egypt and the Internet work-arounds discussion that Egpt sparked
> are in my view historic.
>
> I have no "pride of authorship" on the top level tables, deferring to all of
> you as to whether that structure serves the group.

I think they definitely work for now. See no need to change unless the
need becomes really apparent.

> I do have in the back of
> my mind business plans and funding proposals I have written, the
> organization is intended to cover the big three:  people, things, and money.
>
> Connexus Sumus Unum.....if anyone knows a Latin wizard, would be glad to
> have this checked, intent is "Connected, We Are One"
>

--

Samuel Rose

unread,
Feb 26, 2011, 9:11:00 PM2/26/11
to building-a-distributed...@googlegroups.com, Aaron Huslage, Eugen Leitl

I can understand that it may be that TCP/IP wireless mesh networks are
not a drop-in replacement for the existing wired and controlled
internet we have now.

However, it seems to me that freedombox in particular is a good
starting place for immediate further exploration. I think we should
focus here are where the first best places are to start in evolving
the next internet(s).

*If* TCP/IP wireless mesh networks afford some level of p2p internet
for participants (albeit perhaps not for people in places like Libya,
Iran, etc due to discoverability) then it seems to me that projects
like freedombox warrant further investment and development, if it is
plausible that they can be further developed and refined to eliminate
the need for upstream backhauls, etc etc

If this is *not* possible, then it'd be good to have the background
material on why, plus to at least start answering the question:

*If* TCP/IP wireless mesh networks are *not* the best first place to
start building towards a distributed/decentralized internet, then what
is the best place for us to immediately put our energies towards?

> --
> Aaron Huslage -  +1-919-600-1712
> http://blog.hact.net
> IM: AIM - ahuslage; Yahoo - ahuslage; MSN - hus...@gmail.com; GTalk -
> hus...@gmail.com
>

--

Robert Steele

unread,
Feb 26, 2011, 9:19:06 PM2/26/11
to building-a-distributed...@googlegroups.com
I see HUGE value is a special section:

What Does NOT Work.....

On Sat, Feb 26, 2011 at 8:45 PM, Aaron Huslage <hus...@gmail.com> wrote:

Aaron Huslage

unread,
Feb 26, 2011, 9:20:08 PM2/26/11
to building-a-distributed...@googlegroups.com, Robert Steele
excellent idea. the proof is anecdotal for me. there are still a lot of true believers in mesh, so something could still occur (or have occurred) that i'm unaware of.

Samuel Rose

unread,
Feb 26, 2011, 9:31:07 PM2/26/11
to building-a-distributed...@googlegroups.com, Aaron Huslage, Robert Steele, Paul B. Hartzog, Richard Adler

Can you discuss the problems/objections you've had to TCP/IP mesh
network thus far?

I am wondering, for instance, if there's any recent work done that can
use small world network nearest neighbor models. Here's a really old
page I dug up from archives when I was researching sensor networks a
long time ago that gives some context about what I am thinking here:

http://www.sensorsmag.com/networking-communications/wireless-sensor/the-realities-dealing-with-wireless-mesh-networks-774

The point being here that there may be some models from complex
systems theory that can address the various problems with largeness,
density, congestion, etc by adopting simple rule sets based on various
conditions among the agents acting as routers in the system at any
given time.

Eugen Leitl

unread,
Feb 27, 2011, 4:51:36 AM2/27/11
to Aaron Huslage, building-a-distributed...@googlegroups.com
On Sat, Feb 26, 2011 at 08:45:48PM -0500, Aaron Huslage wrote:

> From what I can tell, they could simply co-opt much of the work that was
> done on OLPC's server and client and spit out a linux distro. Not to mention
> the DD-WRT work. We have this stuff pre-baked and throwing money at the same
> solution over and over when it just doesn't work is not all that useful to
> me.

I haven't joined freedom box project yet, so it's a guess, but I presume
it's pretty much the plan, more or less. It would be quite enough to
have a standard distribution packaged with many useful packages in a sane
preconfigured default, and offer out-of-the-box appliances based on that. This
will boost existing, functioning networks by two or three orders of
magnitude, and give people easy access to them. This can become a point of
departure for more ambitious projects, which without such priming
would be stillborn.



> Wireless Mesh TCP/IP networks simply are not usable, scalable or connectable
> enough to work. We've been trying this for close to 10 years on the Wi-Fi

I agree that conventional wireless meshes are not scalable, and that
they offer only a tiny fraction of a fiber's throughput. However,
this doesn't mean that it's not possible to build scalable wireless
meshes, that the local loop bandwidth won't go beyond GBit, and
that you won't start deploying LoS laser of fiber once the network
is dense enough.

The idea is getting the network dense enough.

> incarnation and it's just beating our heads against 20 walls over and over
> again. They ALL require some upstream carrier to be a backhaul to the

There's nothing wrong with currently using the Internet to emulate long-range
connectivity.

> internet, none of them distribute outbound routes properly, they all get
> heavily congested at a certain point with routing traffic and metadata.

That's the nice point of geographic routing: each node needs to know
only a very small, local part of the global network.



> Not to mention that wireless mesh networks are incredibly easy to find and
> are totally inappropriate for sending into surveillance states like Libya or

That's a different threat model. Making networks functional yet
stealthy is hard.

> Iran...as I've seen get funded lately. The name Wireless Mesh on a grant
> application is damn near guaranteed to make you attractive. This is
> dangerous and stupid and people will get killed. That makes me sad.

I agree that's a stupid idea.


> I'm not saying I have an answer here. I just know what the solution isn't.
> These networks aren't p2p and they don't work.

P2P or grassrouting, or whatever you call it, are just an idea of
self-organizing networks without a central authority.

Eugen Leitl

unread,
Feb 27, 2011, 5:14:23 AM2/27/11
to building-a-distributed...@googlegroups.com
On Sat, Feb 26, 2011 at 09:31:07PM -0500, Samuel Rose wrote:
> On Sat, Feb 26, 2011 at 9:20 PM, Aaron Huslage <hus...@gmail.com> wrote:
> > excellent idea. the proof is anecdotal for me. there are still a lot of true
> > believers in mesh, so something could still occur (or have occurred) that
> > i'm unaware of.
> >
>
> Can you discuss the problems/objections you've had to TCP/IP mesh
> network thus far?
>
> I am wondering, for instance, if there's any recent work done that can
> use small world network nearest neighbor models. Here's a really old
> page I dug up from archives when I was researching sensor networks a
> long time ago that gives some context about what I am thinking here:
>
> http://www.sensorsmag.com/networking-communications/wireless-sensor/the-realities-dealing-with-wireless-mesh-networks-774
>
> The point being here that there may be some models from complex
> systems theory that can address the various problems with largeness,
> density, congestion, etc by adopting simple rule sets based on various
> conditions among the agents acting as routers in the system at any
> given time.

I will venture into kook country, and offer what little original
research I did, some 15 years ago.

I wanted to generalize Bresenham/raycasting like routing for higher
lattice dimensions. Reason: blocking is arbitrarily improbable for
higher dimensions.

I started with n-hypercubes, mapping/projecting nodes and connections to
squares and cubes in 3d space with 2^n by the side. The result was a locally
connected mesh with decaying mesh density, where each link went
twice as far in orthogonal, alternating directions. When plotting
the connectivity matrix of n-cubes it is immediately obvious that
is a fractal, each n-cube being composed of n-1 cubes, interconnected.

If you orthogonalize the connectivity table, you arrive at a hyperlattice,
an orthogonalized hypercube. When mapped to 2d or 3d space, 2^n by
side again the directions no longer alternate. Each link reaches
still twice as far as the previous one.

Routing in such is simply forwarding the packet to a numerical
delta of two node IDs, treated as binary numerical value, repeated
until you arrive at the destination.

While above is still worthwhile to pursue, here's a simpler approximation.
Simulate a high-dimensional lattice, by using existing global networks
as long-distance links. Use the local loop, when node density will
become high enough that adjacent nodes can see each other. GPS (WGS84)
position fixes allow easy derivation of colisionless addresses, and
a proximity metric, deviations from which can be used to build your
local routing table. Incidentally, the /64 local address part of
IPv6 allows enough resolution to store geographic addresses, and that
information will be preserved when delivered by current TCP/IP
protocols.

Devin Balkind

unread,
Feb 27, 2011, 1:36:23 PM2/27/11
to building-a-distributed...@googlegroups.com
On Sat, Feb 26, 2011 at 11:13 AM, Eugen Leitl <eu...@leitl.org> wrote:
On Fri, Feb 25, 2011 at 06:51:13PM -0500, Devin Balkind wrote:
> Eugen.  Enthusiastic students do amazing work so if you'd like to talk
> technical, please post your documentation and we will start following
> your rss at floing and provide you support.

Do you have access to indentured labor, aka C.S. undergrads
and graduates? Ability to run large scale network simulations
is crucial in order to find design flaws and measure
scaling. There's plenty to build and to publish, so academic
ties would be good.


Our organization has had success using 'indentured labor' in the past and we've got good relationships with universities in the New York area - including NYU, Pratt, Columbia, Pace. etc.  College students have tons of free time and are getting VERY interested the free/libre/open source movement (see: Students for Free Culture). 

I can help put together a student engagement plan, but I'm going to need the group to tell me exactly what we need from students.

This might be premature but keep these questions in the back of your mind:
1. what types of tasks can students perform?
2. what qualifications (major, area of study, etc) do students need to tackle these tasks?
3. how many hours does it take to complete each type of task?
4. what type of oversight does each task type require and how can students be held accountable/what artifacts to they produce?
5. what type of person needs to manage students for these tasks?
6. how many tasks need to be done before august/after august?

> If someone doesn't start a Tiddlywiki for this group before tomorrow I
> will start one at tiddlyspot.
>

We're going to start discussing our land's IT system - and it's relationship to the autonomous internet project - on our tiddlywiki here: http://flofarm.tiddlyspot.com/#IT
 
> The other things people have posted on this board have been very
> exciting and look forward to posting more when I get off the road
> tomorrow.
>
> DWB

--
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



--
Devin Balkind
@devinbalkind
vitamindwb.com

Eugen Leitl

unread,
Feb 27, 2011, 3:19:22 PM2/27/11
to building-a-distributed...@googlegroups.com
On Sun, Feb 27, 2011 at 10:36:23AM -0800, Devin Balkind wrote:

> Our organization has had success using 'indentured labor' in the past and
> we've got good relationships with universities in the New York area -
> including NYU, Pratt, Columbia, Pace. etc. College students have tons of
> free time and are getting VERY interested the free/libre/open source
> movement (see: Students for Free Culture).

That is very good to hear.



> I can help put together a student engagement plan, but I'm going to need the
> group to tell me exactly what we need from students.
>
> This might be premature but keep these questions in the back of your mind:
> 1. what types of tasks can students perform?

A literature review of existing systems would be a good first task.
Reporting hands-on experience in using existing systems. Scaling behaviour of
existing systems, using simulators (that will need some hardware
resources for modeling). Investigating efficient routing in node
clouds, which is more a Masters or Ph.D. thesis type of topic.

Characterizing users and designing usability would be a task for
a sociology type of person. Investigating impact of electronic p2p
currencies for economics.

> 2. what qualifications (major, area of study, etc) do students need to
> tackle these tasks?

Computer Science majors would seem a good fit, but in principle
any bright hard science majors would do.

> 3. how many hours does it take to complete each type of task?

A more or less thorough review would take anything from a week to a month.
Exploring scaling behaviour is about a thesis worth of work.

> 4. what type of oversight does each task type require and how can students
> be held accountable/what artifacts to they produce?

That is a good question.

> 5. what type of person needs to manage students for these tasks?

What kind of persons would the institution find acceptable?

> 6. how many tasks need to be done before august/after august?

I don't think we're in any rush at the moment.



> > If someone doesn't start a Tiddlywiki for this group before tomorrow I
> > > will start one at tiddlyspot.
> > >
> >
>
> We're going to start discussing our land's IT system - and it's relationship
> to the autonomous internet project - on our tiddlywiki here:
> http://flofarm.tiddlyspot.com/#IT

--

Paul B. Hartzog

unread,
Feb 27, 2011, 4:36:15 PM2/27/11
to building-a-distributed...@googlegroups.com
Eugen has an interesting solution.
To avoid overlapping you could use either a physical or temporal "location"
as an anchor for addressing schemas.
Einstein would argue there is only a timespace schema,
but he would probably also point out that they are all relative. :-D

All you need to effectively "solve" routing issues in flat networks is

1) ubiquity
2) a definition for 'neighbor'

The win though is neither of these.
The win is that AFTER you have defined various addressing schemas,
and produced hardware that uses them, you

1) do not force any node on the network to "understand" the entire network
2) avoid having only one schema on any given node

To solve the problem of provision,
you simply apply a "commons" solution:
every node must provide more utility than it consumes.
As a consequence every new node on the network
makes the network as a whole run better.
(this is not hard given that we all sleep at night,
i.e. our unused devices can provide for others)

Basically,
you get the best performance when
1) there are MANY definitions of "neighbor" that result in networks via links
2) each node has multiple neighbors in EACH network

As always,
plurality, diversity, functioning on multiple axes all the time....

-p

--

Aaron Huslage

unread,
Feb 27, 2011, 8:04:10 PM2/27/11
to building-a-distributed...@googlegroups.com, Paul B. Hartzog
Isn't this really just the way the Internet works now, just taken down a notch? You still need routers-of-routers to know where these things are...physical location is largely irrelevant in network topologies, or maybe SHOULD be irrelevant. Encoding location information into addressing structures reveals a lot about a user, we see that already with GeoIP-type solutions.

How is a solution that makes the users less secure in favor of easy routing a good solution?

Eugen Leitl

unread,
Feb 28, 2011, 8:22:05 AM2/28/11
to building-a-distributed...@googlegroups.com
On Sun, Feb 27, 2011 at 08:04:10PM -0500, Aaron Huslage wrote:
> Isn't this really just the way the Internet works now, just taken down a
> notch? You still need routers-of-routers to know where these things

There's no such thing. Routers are routers.

> are...physical location is largely irrelevant in network topologies, or

In our relativistic universe, and where distances translate in infrastructure
and earthworks, topology must eventually follow geography. For latency
and economy reasons alone.

NANOG would disagree, but they're extremely invested in centralist mindset.
I think they're going to be surprised.

What I think is eventually coming is cut-through photonic routing (which
is basically a minimal decoration of raw physics of this universe), which
is incompatible with today's most assumptions. Particularly, a 300 m
distance is a 1500 Byte FIFO at 10 GBit/s, and cut-through means the
routing decision must be made within ~ns, potentially using fiber
loops or slow light medium for local buffer -- which means purely
photonic gates, not very many of them, and minimal packer headers
with optimal layouts.

This is just an aside. This is of no concern to us at the moment.

> maybe SHOULD be irrelevant. Encoding location information into addressing
> structures reveals a lot about a user, we see that already with GeoIP-type

Anonymization is a higher layer problem. MIX networks for low-latency,
PIR for high-latency protocols. The advantage of decentral routing is
that global adversaries cannot monitor everything.

> solutions.
>
> How is a solution that makes the users less secure in favor of easy routing
> a good solution?

Onions are like ogers. They have layers.

Reply all
Reply to author
Forward
0 new messages