Fwd: I have the killer application for you....

26 views
Skip to first unread message

Mark Janssen

unread,
Mar 9, 2012, 11:09:18 PM3/9/12
to building-a-distributed...@googlegroups.com
Hello,

I'm forwarding a message I sent to the Internet2 group that I thought might be relevant here on the next-net list.  Of relevance I think would be a discussion about re-vamping the OSI model to accomidate a content-centric, p2p internet.

cheers,
Mark


---------- Forwarded message ----------
From: Mark Janssen <dreamin...@gmail.com>
Date: Fri, Mar 9, 2012 at 9:06 PM
Subject: I have the killer application for you....
To: net...@internet2.edu


Hello,

I'm Mark Janssen and have been working on a information-science project called pangaia.  I think I have an application which can create a true Internet 2.0 that will excite the masses and create a new economy.  I'm sure none of this is anything new, but I hope you'll bear with me.

The idea is to create a content-centric internet.  But this requires a re-vamping of the traditional OSI model.  Instead of the monolithic 7 layer model, the top three layers get inverted and set orthogonal to the physical layers.  The presentation layer becomes a 3-dimensional model organized around a unified epistemic model of information.

But first, let me back up a bit.  The internet has gotten *balkanized* -- the billions of users now spend their time a just a handful of sites.  I argue that this is because of its current topology that was engineered around servers, and hence DNS.  The current interaction model with the internet is all client-server based, yet most content being generated now, is not only transient and mobile on individual machines, but also generally wanting *trying *to be peer-2-peer.  This is the natural way we interact with one another. 

In order to do make a p2p content layer for the internet, the Session layer needs to be at the top and manages the user's view into the next layer:  the Presentation layer which, as I stated, should be a 3-dimensional view of the internet content.   Much like the TCP/IP and the first Internet unified the various networks on the globe, pangaia (tentative name) unifies the *data models*.   Users can create nodes, group nodes into projects, and determine at what level of the network they wish to share or publish their personal content.  Social networking in the p2p model becomes organic and natural.  The data model includes the ability to "vote", as well as rank users, so content can get organized in a meaningful way.  A visual "physics" then creates galaxies of linked content that users can navigate as if they're flying through space.

Think wikipedia in 3-d, plus per-revision voting, plus user ranking.  Instead of the mostly 2-d presentation of HTML with an underlying 3-d logic, you put the 3-d logic near the user and let the sophistication of the visual cortex do the relationship-making.

In any case, this is the final layer which I call the "Content layer", rather than the Application layer, because there is now only the "one application", and it can manage and scale all the different types of interactions on the internet, beautifully.

Anyway, I'm sorry if my wording is a bit bombastic, I've been chomping at the bit trying to drum up interest.  I have a nearly complete specification and an implementation could be made in visual-Python in less than a month.  There's more information at pangaia.sf.net.  A very crude 3-d interaction model is on github.  Please let me know if there's any interest or forward to those who are more inclined to such dialog.

Thanks for your time,

Mark
Santa Fe, NM

Poor Richard

unread,
May 27, 2012, 10:09:11 PM5/27/12
to building-a-distributed...@googlegroups.com
Mark, I am no longer involved in software development -- I just wanted to give you some positive feedback. I don't know if we should get rid of the application layer yet, maybe just change it to the "apps" layer for now. :) But a content layer makes sense to me.

I have been arguing for somebody like the Linux community to do an appliance-like "Peer Box" that provides all the p2p apps needed for social networking, voting, trust/reputation metrics, content collaboration and management, workflow, complementary currency, crowd funding, etc.

PR

Bryce Lynch

unread,
May 27, 2012, 10:54:27 PM5/27/12
to building-a-distributed...@googlegroups.com
On Sun, May 27, 2012 at 10:09 PM, Poor Richard <poor.r...@gmail.com> wrote:

> I have been arguing for somebody like the Linux community to do an
> appliance-like "Peer Box" that provides all the p2p apps needed for social
> networking, voting, trust/reputation metrics, content collaboration and
> management, workflow, complementary currency, crowd funding, etc.

Isn't that what the FreedomBox is supposed to be?

--
The Doctor [412/724/301/703] [ZS]
https://drwho.virtadpt.net/
"I am everywhere."

Poor Richard

unread,
May 27, 2012, 11:12:06 PM5/27/12
to building-a-distributed...@googlegroups.com
Didn't I mention FreedomBox? That's where I got the PeerBox idea, but the FB function is only security, it doesn't have the p2p apps package I'm talking about. It might be an ideal host for them, though.

Isaac Wilder

unread,
May 27, 2012, 11:13:30 PM5/27/12
to building-a-distributed...@googlegroups.com
I think you've misunderstood FreedomBox. That's precisely what it's supposed to be.


Isaac Wilder
www.thefnf.org

Poor Richard

unread,
May 27, 2012, 11:37:02 PM5/27/12
to building-a-distributed...@googlegroups.com
Doc,

If the FreedomBox will ship complete with "all the p2p apps needed for social networking, voting, trust/reputation metrics, content collaboration and management, workflow, complementary currency, crowd funding, etc." (not a complete list) its news to me and I followed the listserve for six months.

I'm not talking about something that enables the secure use of such apps, I'm talking about coming complete with the apps themselves pre-installed and integrated.

That can't be what FreedomBox is because those p2p apps don't yet exist as an integrated package, or even as individual apps that are adequate to replace Facebook, Google docs, Google search, Wikipedia, YouTube, Kick-Starter, Liquid Feedback, etc. etc.

The requirement for each app is:

  • world class
  • open source
  • p2p architecture
  • integrated with each of the other apps in the suite


If you think that exists, please show me.

PR

Isaac Wilder

unread,
May 28, 2012, 12:25:30 AM5/28/12
to building-a-distributed...@googlegroups.com
You're right, it won't ship with all that, but I do believe it will have social networking built on top of trust/reputation metrics. That's the core infrastructure of what you're talking about. Once the distributed social graph and peering/discovery protocols are in place, adding applications should be relatively simply. (Assuming we can get them built :-))


imw

Mark Roest

unread,
May 28, 2012, 2:03:13 AM5/28/12
to building-a-distributed...@googlegroups.com
Hello Mark, Isaac, Poor Richard and all. It looks like two different philosophies of programming at work here. Isaac seems to be assuming the market / the crowd will come forward with sufficient applications for each category in due course. Poor Richard is calling for a systems approach, fully integrated, world class, and able to compete with best-of-breed commercial versions of each application -- more of an integrated project approach, like the Linux kernel. I believe Poor Richard is one of the pioneers of computing -- one of the old guard whose collective ideals got marginalized by corporate drives for consumerism, for a while, but are now rallying for common visions of what is possible for computing in service to humanity.

I agree with Poor Richard, and go a step farther. I would like to see these applications integrated into an open source, P2P-capable, digital earth imaging / GIS knowledgebase foundation, whose primary organizing categories are eco-regions (and ecosystems), cultures, and a systematic approach to supporting humans without compromising or killing off other species -- the user interface of my approach to a New Operating Manual for Lifeship Earth. It is based not on western consumer / military culture, but on the ecologically sound ways of meeting human needs that sustained each culture for thousands of years, augmented by the truly useful and non-harmful cream of the technological crop -- developed in the Maker subculture where that makes sense. (It makes perfect sense that the machines and skills which the Maker culture develops for pleasure would often be a perfect fit for empowering cottage industry and bountiful subsistence agriculture to bring prosperity to rural village life -- and to urban neighborhoods as well.) That is very different from the global industrial lifestyle; it is local lifestyles, in harmony with the natural places in which they occur.

Here is where the importance of integration comes in: each individual in each place, and the local village cooperative and its sub-units, are able to input their perception of reality into the information for their place -- and it can, on demand, appear for any substantially similar place on the planet as one approach to whatever the topic is. This information can include sensor data or human observations in nature and agriculture; statements of market demand, offers of supply, coordination of supply, and transactions entered into; development plans; disasters and preventive and mitigating disaster plans; and educational modules made by teachers, scientists, griots, shamans and other healers, museums, technology companies and inventors, 'regular folks', and students in constructivist educational settings. 

All of this information needs to be able to be aggregated to successively larger scale places, up to and including global. It can create accurate, actionable meta-information about how much capacity is needed in a given industry or crop in any place, making it possible to greatly reduce waste and duplication, as well as to plan migration paths for the paradigm shift, limiting disruption of vital services. It can also create the equivalent of large painted targets for those of us who are bringing new technologies to market, enabling us to jump over barriers to entry in each industry by connecting with niche markets, whether categorical or place-based. (This has significant implications for the use of money.) To bring even more power to these benefits, we can create regional environmental-econometric input-output analysis databases, which are essentially fractal, representing physical transformations as well as monetary transfers, and able to both integrate their results into larger region sets, and decompose them for smaller places or niches. This becomes the ultimate economic power-tool for the people -- for lifestyles of community and collaboration gone global.

Figuring out how to make this work, and implementing it, is going to take the collective skills of many of the most skilled information systems architects. Once it exists, it will also provide fertile ground for custom tailoring native applications to meet the specific needs of organizations and individuals -- and it will inevitably create a new industry of people integrating it with legacy applications.

Mark, your work might possibly be able to be part of a valuable proof-of-concept and sales tool, to help get support for the larger mission, and give people a great tool in the meantime. It might also be a social module of the whole, whether the particular programming languages you used are retained, or the structure is recompiled using the programming language and architecture chosen for the knowledgebase.

Regards,

Mark Roest

Poor Richard

unread,
May 28, 2012, 4:44:42 AM5/28/12
to building-a-distributed...@googlegroups.com
Mark, I agree with the importance of gps and Google Earth (or equivalent) functionality. In the beginning it will be necessary to have interfaces/connectors to various proprietary client-server applications like Google's until they can be re-engineered in p2p open source versions. As for a database engine to support your vision, that has been discussed here before and I think there are some possible p2p candidates. I'm glad you reminded me of all that modelling stuff -- I'm about to take a stab at a spec for...

PeerBox -- Peer-to-Peer Everything (modular application suite, developer's tool kit, and security appliance in a box).

PR

Melvin Carvalho

unread,
May 28, 2012, 7:26:51 AM5/28/12
to building-a-distributed...@googlegroups.com
On 10 March 2012 05:09, Mark Janssen <dreamin...@gmail.com> wrote:
Hello,

I'm forwarding a message I sent to the Internet2 group that I thought might be relevant here on the next-net list.  Of relevance I think would be a discussion about re-vamping the OSI model to accomidate a content-centric, p2p internet.

3d visualizations ... anything similar to Ted Nelson's zig zag data?

http://www.youtube.com/watch?v=WEj9vqVvHPc

I like your idea of voting and reputation.
 

Melvin Carvalho

unread,
May 28, 2012, 7:41:56 AM5/28/12
to building-a-distributed...@googlegroups.com
On 28 May 2012 05:37, Poor Richard <poor.r...@gmail.com> wrote:
Doc,

If the FreedomBox will ship complete with "all the p2p apps needed for social networking, voting, trust/reputation metrics, content collaboration and management, workflow, complementary currency, crowd funding, etc." (not a complete list) its news to me and I followed the listserve for six months.

I'm not talking about something that enables the secure use of such apps, I'm talking about coming complete with the apps themselves pre-installed and integrated.

That can't be what FreedomBox is because those p2p apps don't yet exist as an integrated package, or even as individual apps that are adequate to replace Facebook, Google docs, Google search, Wikipedia, YouTube, Kick-Starter, Liquid Feedback, etc. etc.

The requirement for each app is:

  • world class
  • open source
  • p2p architecture
  • integrated with each of the other apps in the suite

The web is already a peer to peer architecture.  I think the linked data ecosystem is the ideal place to incubate and grow these concepts.  The big question is whether the system attempts to replace DNS from the get-go, or bootstrap it. 

Even on a 10 year time line with billions of capital, it would be quite a challenge to create any significant challenger to DNS.

Bootstapping the existing web architecture and allowing upgrade paths vis the "universal" nature of the URI, I think is the way to go.
 

Mark Janssen

unread,
May 28, 2012, 11:02:56 AM5/28/12
to building-a-distributed...@googlegroups.com
On Mon, May 28, 2012 at 6:26 AM, Melvin Carvalho <melvinc...@gmail.com> wrote:
On 10 March 2012 05:09, Mark Janssen <dreamin...@gmail.com> wrote:
Hello,

I'm forwarding a message I sent to the Internet2 group that I thought might be relevant here on the next-net list.  Of relevance I think would be a discussion about re-vamping the OSI model to accomidate a content-centric, p2p internet.

3d visualizations ... anything similar to Ted Nelson's zig zag data?

It's very much like project Xanadu.  The difference is that the pangaia project starts from the basis of information theory (the bit) and epistemology where the Xanadu project really started at the idea of the "document" and a cruder idea of "knowledge".  Without a social basis, knowledge has little basis.

Thanks for the response!

Mark

Mark Janssen

unread,
May 28, 2012, 11:16:16 AM5/28/12
to building-a-distributed...@googlegroups.com
On Sun, May 27, 2012 at 9:09 PM, Poor Richard <poor.r...@gmail.com> wrote:
Mark, I am no longer involved in software development -- I just wanted to give you some positive feedback. I don't know if we should get rid of the application layer yet, maybe just change it to the "apps" layer for now. :) But a content layer makes sense to me.

Yeah, it is a bit radical, but then if someone would have suggested that all the various networks and BBS would be unified 10 years from then, probably no one would have believed it either...

Thanks for the feedback!

Mark

FC Fisher

unread,
May 29, 2012, 2:38:36 AM5/29/12
to The Next Net
Here's a similar approach that is a bit further along: http://qualityofid.com/
The alpha peer is due out in a couple of weeks. The infrastructure
that turns the network into the computer (processing and data control)
is already in use by HP and Microsoft.
We'd welcome additional participation, but also value that at this
stage of p2p development multiple approaches will increase the
likelihood of a viable solution.

Fred

On May 28, 8:16 am, Mark Janssen <dreamingforw...@gmail.com> wrote:
Reply all
Reply to author
Forward
0 new messages