by Geert Lovink, founding director of the Institute of Network Cultures, June 6, 2012
I was asked by an Italian newspaper to submit 96 words on the future of internet. Here they are:
Instead of further going down the corporate lane of Microsoft, Apple, Amazon, Google and Facebook, I propose to go back to the original architecture of Internet as public infrastructure with decentralized nodes. It may be romantic to insist on the distributed nature of networks but it is a necessary political demand. Net criticism is a toothless project without a utopian dimension. Even if internet itself had a military origin in the Cold War, and is now dominated by equally destructive force of greedy venture capitalists, backed up by libertarian gurus. Let’s rethink the public sphere: another internet is possible!
Efficiency is going to dictate that some peers are going to serve different distribution roles in the network. How and why they do this is critical to acceptance of any system.
Each peer class might have several sub-classifications:
On another topic, I suggest that all the PeerPoint applications (email, social chat, project spaces, wiki pages, etc.) need to have common elements in the user interfaces to set and override default privacy and security settings at a fine level of granularity (e.g. individual emails, files, documents, comments, links, etc.).
User-configured Security settings (by content categories and by individual data items) might include:
Finally, the extreme nature of some of the privacy
and security options mentioned above raises the danger of 1) provoking
government reaction against otherwise legal and benign private network
activity and 2) potentially attracting the use/exploitation of PeerPoint
technology by anti-social actors, which would exacerbate danger #1. For
these reasons PeerPoint networks would need mechanisms for policing
themselves against anti-social uses, perhaps including methods for
segregating high and low risk activity on separate networks, and perhaps
some way of trying to satisfy the minimum national security
requirements of liberal democratic states as far as possible.
I've shared PeerPoint Next Net group members: I've shared a Google Doc called "PeerPoint" with the group. Members have edit permission but please follow the two simple procedures mentioned at the top of the document.
Click to open:
Google Docs makes it easy to create, store and share online documents, spreadsheets and presentations.
In order to use these you need to
The PeerPoint repository currently has just a readme file and the wiki has just a text version of the PeerPoint Specifications with most of the links stripped out of it.
The objective is to get all the PeerPoint documents and activity shifted over there as soon as possible. But I am new at GitHub and no wiki-master either, so I can use all the help I can get.
I assume it will generally be best to put code in the repository and docs in the wiki. We don't have any code and might not for a while, although there is plenty of relevant open source code we can copy to our repository as needed. The actual PeerPoint design specification should probably go in the repository so we can easily fork it and make use of the version control features.
The wiki can hold all the design specifications and documentation gathered from similar or related open p2p projects. There is an Open Design project on GitHub that may contain material relevant to the PeerPoint Open Design process.
If there is enough interest I'll set up a dedicated PeerPoint Google Group and/or mailing list.
If anyone is willing to help set up the GitHub site please let me know.
Cheers,
PR
BTW as far as I can tell, there is no online forum dedicated to the general topic of p2p application software design for free/open source software designers, developers, engineers, etc. Discussions of p2p architecture tend to be found mostly in the discussions of specific p2p networking projects like freenet and FreedomBox and they tend to apply to the lower-level network routing and file-sharing issues instead of the higher application software level for real-time, interactive collaboration. If anyone knows of a general p2p application architecture group or forum, please let me know.
Currently, distributed CVS systems (notably GIT) are a peer-to-peer
application that's very successful. And then there's torrent.
Re. Wikipedia and Linux: Both build on rather rich ecosystems of prior
art, and development communities. Wikis came first, in various flavors,
wikipedia. The world wide web virtual library predated wikipedia by
decades. It took a lot of experimentation, plus vision and drive before
wikipedia coalesced. Linux builds on the legacy of Unix and lets not
forget all the stuff done by Stallman and the FSF (e.g, all the gnu
tools and gnu userspace without which Linux would be useless).
The problem with Poor Richard's proposal is severalfold:
1. lack of specificity - "let's build a P2P application suite" is a
pretty content free statement - the PeerPoint "specification" really
isn't anything resembling much of a specification
2. lack of vision - Ted Nelson's "Dream Machine's" book and Xanadu
concept never really led to any widely used code, but they sure inspired
a lot of people - precisely because Nelson laid out a vision of new
capabilities -- other people went on to implement some of those ideas
3. lack of understanding about how technology evolves - open source or
not - it involves an ecosystem and a process of incremental development,
punctuated by flashes of insight
4. lack of anything motivating - is there really that much
dissatisfaction with FaceBook? I mean, an awful lot of people use it,
and there are alternatives for old reprobates like me who find FaceBook
pretty insipid and useless (except for keeping track of my kid). For
real work, email lists have been around an awfully long time, and don't
seem to be going away anytime soon. The folks who REALLY worry about
privacy/secrecy are using VPNs, encryption, darknets, etc.
"Let's build an application suite" is a mantra that leads to things like
MS Office. "Let's build an open source application suite" leads to
OpenOffice and LibreOffice - which are essentially not-as-good
imitations of MS Office. I simply don't see anything in the PeerPoint
proposal that goes beyond "let's do x, y, and z, but let's do it
peer-to-peer."
Now if Poor Richard were to start developing stuff, that actually did
useful and interesting things, I expect it would attract a community.
But just saying "I think people should do <this>," where <this> is
generally uninspired and uninteresting, is not a recipe for anything.
</end rant>
Paul Hughes wrote:
> Mark,
>
> Wikipedia and Linux are two prime examples of substantive creations greater than Poor Richard's proposal that were done almost entirely by volunteers with no desire for compensation other than the joy of creating something awesome.
>
> I believe the time is ripe for PeerPoint or something equivalent. In fact it's long overdue. Because of the growing dissatisfaction with Facebook, and efforts to control and censor the net, there are more people wanting, and willing to build something like this than ever before.
>
> Paul
>
>
>
> On Jun 18, 2012, at 20:15, Mark Janssen-Rosenbluth <dreamin...@gmail.com> wrote:
>
>> Richard, there is very little such application suites on the "drawing board" because we're way ahead of you. These conversations have been going on for over a decade. You're excited to get started, so you start afresh, but many of us have heard it all before. That there isn't something more by now is a difficult fact to explain. But probably the biggest reason is that after the litigation over p2p file-sharing and such, no one (with financial resources or political clout) has taken the "leap of faith" to "get it done". Such a comprehensive and radical project isn't that far from the process that created the protocols of the whole Internet to begin with... and you know how many years and millions that took....
>>
>> So, if you want this to happen, I suggest you solve the political and/or financing problem.... don't go rei-inventing the wheel.
>>
>> mark j
--
In theory, there is no difference between theory and practice.
In practice, there is. .... Yogi Berra
On Tue, Jun 19, 2012 at 9:15 PM, Poor Richard wrote:
> But I don't agree on one technicality: I don't see that P2P per se advances
> co-intelligence that much.
Right, you mean, just throwing in "the ultimate P2P tools" into a
group doesn't necessarily turn them into a
coherent-whole-organism-thingy?
Co-intelligence -- as far as I know, from reading and interpreting Tom
Atlee's stuff and from joining that to some stuff I've seen -- seems
to depend heavily on something "genuine" that we carry "inside"...
... This seems to be related with
"trust": I "trust" this group, so I'll show more of my imperfect human
self. This allows both for the group to identify what are the common,
real, deep and imperfect human longings that we all share, and for
people to heal whatever wounds they might be carrying...
That may or may not happen any better if you have better means of
electronic information exchange. I would think that the above process
would instead identify the core human longings, which in turn would be
served by a (very) different set of communication tools and
technologies than the ones we have today.
What I'm reaching for is that P2P architecture is a solution to
something which is a greater "sum" of all the individual P2P
motivations you can find. If that makes any sense at all, then, in
addition I suspect that that something is a deeper, more basic issue;
it is at a different, deeper level than all of the individual P2P
motivations, and that discussing at that level is much, much harder
and painful than discussing at the individual motivation level, which
in turn is harder than discussing technicalities (which are completely
external to us and that we can all easily perceive whether they either
work or don't work, so it's easier to separate what's false from
what's true).
> That said, however, I am not above selling P2P with something more positive
> and inspiring than ownership and security.
That stuff is closer to the "discussion" I want to have, but, I'm
starting to suspect mailing lists (asynchronous, group, written
communication) won't be of any help. And I don't have the skill to
handle a complex, inevitably "combative" discussion on these topics.
Me fail.
What I'm reaching for is that P2P architecture is a solution to
something which is a greater "sum" of all the individual P2P
motivations you can find. If that makes any sense at all, then, in
addition I suspect that that something is a deeper, more basic issue;
it is at a different, deeper level than all of the individual P2P
motivations, and that discussing at that level is much, much harder
and painful than discussing at the individual motivation level, which
in turn is harder than discussing technicalities (which are completely
external to us and that we can all easily perceive whether they either
work or don't work, so it's easier to separate what's false from
what's true).
If you're genuine, then I urge you to evaluate the pangaia project. It is a comprehensive project *designed* to unify all the p2p individual desires into a unified theoretical framework as well as new currency model. Because of this, it goes beyond the concept of disparate applications quasi-unified in an "application suite", in the same way the various networking modalities prior to the Internet (compuserve, AOL, tymenet, etc) all got unified under TCP/iP. Check it out: <pangaia.sf.net>
"I am not above selling P2P with something more positive and inspiring than ownership and security."
Certainly PR, no problem at all to quote.Public domain.Sepp
I'm trying to figure out what the big vision is, and what's the
motivation for folks to use it, much less develop it (there are lots of
people who've developed and use peer-to-peer applications and underlying
peer-to-peer technology infrastructure - but the stuff hasn't caught on
the way facebook or twitter have - something is missing).
The PeerPoint Open Design Specification is not meant to replace or supersede existing software and technology development efforts. It is intended to help coordinate the work of the floss/hacker/p2p community towards a future point of convergence and interoperability. It is essentially a statement of user requirements and guidance on preferred technical solution sets. It describes what the progressive user community needs from the technical community in order to prevail in the social, political, and economic struggles that lie ahead.
The Arab Spring, the Occupy Movement, Los Indignados, and similar uprisings around the world demonstrate that a new, open society and open democracy is struggling to rise from the bottom up. But the internet has been colonized by giant corporations whose business models are based on central servers, proprietary technology, surveillance and censorship, and unilateral terms of service.
At best, we might be
able to identify a list of things that work better than others, for some
applications.
Now, if you have a clear vision and approach to the contrary, let's hear
it - but with specifics more like those in the gnunet or git
documentation
- or for that matter, something that lays out something
distinct from the freenet project (https://freenetproject.org/) - which
has been around a long time, and DELIVERS what it promises:
"Freenet is free software which lets you anonymously share files, browse
and publish "freesites" (web sites accessible only through Freenet) and
chat on forums, without fear of censorship. Freenet is decentralised to
make it less vulnerable to attack, and if used in "darknet" mode, where
users only connect to their friends, is very difficult to detect."
One of these things is not like the other: 1) freenet 2) PeerPoint.
Which is it?
It might help to have some discussion of why Freenet, and other things
like it, have not caught on. It's certainly not a technical issue -
it's something else, but what? (By the way, I think Freenet is pretty
cool - but not very useful for interacting with the vast majority of
people I deal with daily - they're all using email, Facebook, and
twitter, and preparing documents in Word an PowerPoint. Such is life.)
----
So no, I can't address the document section-by-section, because my
general reaction is that you need to rethink your problem statement,
goals, and approach, and then go back to the drawing board.
Miles
Let me be more precise. If YOU, PERSONALLY, have a clear vision and
approach, let's hear it. If you don't, what's your value add to the
process?
>
> - or for that matter, something that lays out something
> distinct from the freenet project (https://freenetproject.org/) -
> which
> has been around a long time, and DELIVERS what it promises:
> "Freenet is free software which lets you anonymously share files,
> browse
> and publish "freesites" (web sites accessible only through
> Freenet) and
> chat on forums, without fear of censorship. Freenet is
> decentralised to
> make it less vulnerable to attack, and if used in "darknet" mode,
> where
> users only connect to their friends, is very difficult to detect."
>
>
> My initial notes on freenet for my comparison matrix were: "Category:
> intranet/VPN. Closed intranet--no www access. Distributed cache, every
> node allocates space for anonymous encrypted content. No manual purge."
>
> One of these things is not like the other: 1) freenet 2) PeerPoint.
>
Well yeah - FreeNet is a peer-to-peer application suite, well defined,
operational, with a user and developer community.
PeerPoint is undefined
and vapor.
> I agree it would be useful to discuss obstacles to the adoption of
> freenet and other existing projects. What are your thought about it?
>
Sure.. people go with what's easy and free and cool (for some
definitions of cool). Only those who have a compelling need (folks
looking for free media and/or porn, military, revolutionaries,
criminals) or are driven by philosophical biases (cyberpunks) do
anything else.
Fair enough, however.... nobody is just a "nobody" - you're hiding
behind Ben Franklin's pseudonym. Do you have a name, a job, an
education, life experience, context, professional skills? Where are your
ideas coming from? What research lies behind them? .....
Miles,
I'm sorry you don't want to offer anything more constructive than "write a science fiction novel" and "where there is a need to bypass central control, people seem
to find ways."
Does anyone know of an existing open semantic ontology/taxonomy that might be adapted to categorize the players/projects/products in the p2p ecosystem?
PR
Does anyone know of an existing open semantic ontology/taxonomy that might be adapted to categorize the players/projects/products in the p2p ecosystem?
PR
mark
Poor Richard wrote:
> I'm disappointed but not surprised that the PeerPoint project is
> viewed with scorn by old-guard digital libertarians and anarchists.
Please don't count me as either of those, as I am neither.
PeerPoint strikes me as a classic case of analyze,
analyze, analyze, then design the one true solution, by committee.
One of the clear lessons of the past decade or so, is that developing
software in today's world is a lot more like gardening, or maybe
farming, than design and manufacturing. Define simple interfaces,
create a climate, fertilize, cut weeds, do some selective seeding, trim
stuff, and otherwise stand back.
The clearest example is SOA vs. REST. Huge amounts of time and money
have gone into defining SOA architectures, but noboday actually uses
them. On the other hand, we have lots of mashups. There are lessons to
be learned about developing software ecosystems.
My sense is that PeerPoint is based on a set of premises and approaches
that simply don't work anymore.
The Web itself, of course, was designed to be peer to peer.
The DBpedia Ontology is a shallow, cross-domain ontology, which has been manually created based on the most commonly used infoboxes within Wikipedia. The ontology currently covers over 320 classes which form a subsumption hierarchy and are described by 1,650 different properties.
With the DBpedia 3.5 release, we introduced a public wiki for writing infobox mappings, editing existing ones as well as editing the DBpedia ontology. This allows external contributors to define mappings for the infoboxes they are interested in and to extend the existing DBpedia ontology with additional classes and properties.
Overview of the class hierarchy of the DBpedia Ontology
This website is used for the project on Scaling up Common-Pool Resource (CPR) Theory. It is a semantic wiki, which is a wiki that holds data. The data in this site describe the values of variables for a set of example social-ecological systems (cases). These data are held in the form of instances of categories. The categories are shown in the sidebar to the left under "Data". For more information on these categories and adding instances of each of them, click on the "Methods" link to the left. What follows here is a brief description of the contents of this site, the outline of which is given in the sidebar to the left.
The following 5 pages are in this category, out of 5 total.
I am copying Jack Park, who may be able to offer assistance here.
Poor Richard wrote:Well, for what it's worth, that's what I'm working on these days. HTML5 is just about up to the job, but there are serious limits to what current browsers can do vis-a-vis P2P protocol exchanges. That should change, but for now, you pretty much have to use some kind of work-around, either a plug-in, Java applet, or a protocol proxy (either local or on the net).
On Saturday, June 23, 2012 1:47:52 PM UTC-5, melvincarvalho wrote:
The Web itself, of course, was designed to be peer to peer.
Right, and a lot of PeerPoint could be done just in the browser (I know this was not necessarily your suggestion--I'm just brainstorming).
Opera Unite started going that way and there are/were a few FireFox plugins for social and file sharing.
On Sun, Jun 24, 2012 at 12:59 PM, Miles Fidelman <mfid...@meetinghouse.net> wrote:
Poor Richard wrote:Well, for what it's worth, that's what I'm working on these days. HTML5 is just about up to the job, but there are serious limits to what current browsers can do vis-a-vis P2P protocol exchanges. That should change, but for now, you pretty much have to use some kind of work-around, either a plug-in, Java applet, or a protocol proxy (either local or on the net).
On Saturday, June 23, 2012 1:47:52 PM UTC-5, melvincarvalho wrote:
The Web itself, of course, was designed to be peer to peer.
Right, and a lot of PeerPoint could be done just in the browser (I know this was not necessarily your suggestion--I'm just brainstorming).
Opera Unite started going that way and there are/were a few FireFox plugins for social and file sharing.
... a separate app out of the browser (like BitTorrent/Gnutella) will be the only way to go.
It's been done before - a couple of years ago there was an experimental peer-to-peer socnet called Peerscape. It was written in Python and was installed as a Firefox add-on. It worked pretty well but didn't catch on, partially because the Python-in-Firefox plugin isn't maintained anymore, so the rest of the project died. Still, it might be useful to take a look at their code to see how they did it.
---
The Doctor [412/724/301/703]
Web: https://drwho.virtadpt.net/
Sent from a Global Frequency satphone.
Poor Richard wrote:
>
> PeerPoint is not meant as "one true solution" or one true anything.
well.... that's the impression given by:
"a suite of integrated peer-to-peer applications to include (but not
limited to) social networking, real-time project collaboration, content
management, database management, voting, trust/reputation metrics,
complementary currency, crowd funding, etc." that sure sounds like it
and,
"PeerPoint is intended to be much more than a user-owned social
networking platform to replace Facebook, Twitter, etc. It is imagined as
a peer-to-peer (p2p) social collaboration suite, developer’s tool kit,
and security appliance in one cheap plug-n-play box."
and "The PeerPoint is designed to Occupy the Internet."
> It is an attempt to 1) define a certain problem space and 2) catalog a
> related solution space and 3) to do that via an open process.
except that the problem space sounds awfully close to "everything"
> Do you see all open process the same as "by committee"?
>
Not all, but in my experience, it generally ends up that way.
What I've seen work well: Single author or small team gathers some
input, puts a vision on the table, collects feedback, puts together
prototypes, gets more feedback, refines based on feedback, releases
initial version, .... over time, if there is positive reaction and
uptake, involve more people... evolves into a community.
What I've rarely, possibly never seen work well, is: We're going to
through the floor open to input from everyone, at the start, develop a
spec that contains a huge laundry list of requirements, then build to
that spec.
In buildings, there are good architects, bad architects, and middling
architects - but I've yet to see a beautiful and functional building
designed by committee. Design competitions yes, maybe an open process
for framing what's being looked for, an maybe an open judging process -
but design by committee doesn't work.
> I'd prefer to find an existing PeerPoint-like effort and participate
> in that -- but because I haven't found that, PeerPoint is the result.
> PeerPoint can always be folded into another project if a better one is
> discovered.
>
> There are many previous and ongoing efforts to produce requirements
> definitions and solution specifications for particular projects. I
> don't intend PeerPoint to be better than any of those or to replace
> any of them. The aims of PeerPoint are
My basic feedback on this is that I've yet to see something like this
work. You've been out of the field for a decade - in that time, an
awful lot of folks have tried things like this in various fields, and
they've yet to work. I have personal experience of this in the GIS and
military simulation spaces, and have seen failures in a lot more. There
are serious lessons to be learned from those failures, and some
conclusions that can be drawn as to what might work "better" (see below)...
> 1) to define user requirements at both the individual and the social
> scale,
That kind of assumes that people really know what the requirements are.
People don't always know what they want or need, until someone shows
them something new, and which version of something is going to "work" is
never quite clear. (Nobody knew that anybody would use FaceBook until
someone put it online. Nobody really knows why MySpace failed, and
FaceBook is so popular. Makes it kind of hard to write the "FaceBook
killer.")
My personal observation and opinion is that simple tools "work" a lot
better than more complicated ones, and people tend to build more
complicated applications by using simple ones as "building blocks."
In short - requirements are hard or impossible to define in any
meaningful way - what you usually end up with is a laundry list of
incompatible pseudo-requirements that are off the mark from what's
really needed.
What works is to put a vision on the table and see how people react. Or
to get lots of visions on the table, some of which flourish, most of
which don't. (Call it "the market" or "let a thousand flowers bloom"
depending on your political persuasion.)
> 2) to catalog work that falls within the scope of free/open/p2p
> applications and infrastructure, and
Might make a good survey article, but pretty hard to keep up to date.
Besides, we have WikiPedia. Do we need yet another catalogue?
Caveat: there are some examples where cataloging is useful - I
personally find the "mashup matrix" to be a useful tool - but a HUGE
amount of work goes into maintaining that as a highly detailed
collection of interface details.
> 3) to serve as a nexus for collecting information from and providing
> information to interested parties in both the user community and the
> developer community.
Again, lots of people try this, with limited results or benefit. When
users are looking for something, they Google or go to WikiPedia, or
BestBuy, or an Apple Store, or ask friends - until they find something
that is close enough to what they're looking for. Developers are
usually scratching a personal itch (or working for hire), and do their
own research.
> For example, the Foundation for P2P Alternatives has a list of
> projects relating to "p2p infrastructure". There is no effort to
> compare, contrast, or rate the projects according to a uniform set of
> criteria. Wikipedia has a number of pages that compare open source or
> p2p software in specific categories, but the pages use a variety of
> criteria and the pages are not all interlinked. One thing PeerPoint
> collaborators might do is create a Wikipedia page to index the pages
> on p2p software and infrastructure, and define some common criteria
> that might be incorporated into the various comparison matrices.
Cases in point. It might be worth asking WHY those pages are not
adequate. My own opinion:
- they're good enough
- there's not enough demand for something "better"
- it's hard work to keep that kind of thing up to date
- folks who need serious details - be they looking for software, or
developers - are going to do their own research
(Recent example: I've been looking for a video editing package for the
Mac, either free or cheap. It took all of about 10 minutes to go
through this list of open source video editors on WikiPedia, narrow it
down to those that run on the Mac, and then do my own review of the info
on three different web sites.)
> For another example, I assume things relevant to PeerPoint have been
> done or are in progress in the W3C community. I'd like to collect
> information about those things and provide a portal to that activity
> not just for developers but for a wider user community as well.
I guess I keep coming back to.. why? Is there interest?
> Each individual already has her own mental map of the big picture, but
> how can it hurt for us to have an open process and vehicle for
> analyzing and comparing our different versions? That can be done in
> free-form conversation, of course, but it can also be helpful to
> participate in a more structured analysis and formulation.
It takes time and effort, and folks motivated to do so. Discussions
like this happen in a lot of places. Do we need another one? If you
see a personal need - is your effort better expended in improving an
existing resource?
> I may be misunderstanding you Miles, but you seem to have said at one
> time or other that efforts to focus on the big picture are either
> arrogant, stupid, pointless, or impossible.
>
>
> One of the clear lessons of the past decade or so, is that developing
> software in today's world is a lot more like gardening, or maybe
> farming, than design and manufacturing. Define simple interfaces,
> create a climate, fertilize, cut weeds, do some selective seeding,
> trim
> stuff, and otherwise stand back.
>
You are misunderstanding me.
I'm saying that trying to DEFINE the big picture, in detail, in advance
is (to use your words) arrogant, stupid, pointless, and impossible.
Identifying trends, aspects, visions, ... is (maybe) possible and worth
doing (though chaos theory might suggest otherwise).
Fielding's PhD thesis
(http://www.ics.uci.edu/~fielding/pubs/dissertation/top.htm), that
framed the REST architectural model, is a really good and influential
example of taking a big picture look at things. Identifying and
understanding fundamental architectural choices - e.g., SOA vs. RESTful
interfaces - is really important when thinking at the systems level.
(Note, however, that Fielding essentially summarized and codified years
of operational experience in the Internet).
The Unix approach of "everything is a file" is again, a fundamental
architectural choice that effects huge amounts of things.
Raymond's paper on "The Cathedral and the Bazaar" is a critical piece of
work on alternative ways to develop software.
If anything, the world of collaborative software requires:
- better building blocks (vs. big, complicated systems)
- an ecosystem model ("The Cathedral and the Bazaar" as applied to the
evolution of social/collaborative software)
> Even third world small farmers are using cell phones, computers and
> websites to share information on production methods and market
> conditions. That's really all PeerPoint is about.
So.. PeerPoint is about cell phones, computers, websites, ..... - again
pretty broad and amorphous.
> The clearest example is SOA vs. REST. Huge amounts of time and money
> have gone into defining SOA architectures, but noboday actually uses
> them. On the other hand, we have lots of mashups. There are
> lessons to
> be learned about developing software ecosystems.
>
>
> SOA was of, by, and for client-server architecture, not p2p. Software
> as a service was IMO mostly motivated by corporate bean counters
> drooling over a subscription-based or fee-for-service business model.
> They kept the client-server but eventually went to the
> advertising-based financial model.
That's simply wrong. It may be how it's played out, but SOA is about
standard interfaces between software components. And there's a lot more
software out there than subscription-based services sold to consumers.
> Lessons learned have two aspects, historical facts and
> interpretations. Our interpretations don't seem to agree very often.
>
> I would also be more impressed with "lessons learned" if the current
> situation were not so terrible. What is the evidence that we have
> learned very much? Can you point to some projects that demonstrate
> lessons learned?
>
Umm... Linux, Apache, the email ecosystem, chat, the huge universe of
open source software for everything under the sun, the Internet (global
infrastructure that reaches everywhere that we use for everything),
flash mobs, the Arab Spring, crowdsourced responses to natural disasters
(crowd mapping, Ushahidi), the list goes on.......
> The poor adoption rate of technology like freenet should prove that
> there are still lessons to learn. I've learned that problems can't be
> addressed in isolation and solutions can't present potential users a
> dilemma of trade-offs they can't reconcile or the risk of consequences
> they can't foresee. There are only so many single-issue (or small set)
> solutions that users can evaluate, install, and learn before they get
> adoption fatigue. Applications can not be developed as monolithic
> swiss-army-knife application suites, but they need to be able to be
> installed that way and look and feel that way to end users. Linux is a
> pretty good example. Its adoption increased proportionately along with
> distros that offered more apps, integrated them better, and were
> easier to install and use.
Or.. that there isn't a broad need, or a perceived need.
My take is that people aren't adopting FreeNet because they don't see a
need for it.
Good case in point: Napster was incredibly popular - who doesn't like
free music. But... conditions changed, partially onerous legal actions
by the recording industry, but also Apple iTunes and the 99cent
download. There are lots of dirt cheap MP3 players, but people spend a
lot more on iPods.
People use FaceBook, they don't use MySpace, and
my personal opinion is
that creating an alternative open social ecosystem is a complete
crapshoot - because people just don't care.
> PeerPoint is not a spec for a monolithic solution--it is a spec for a
> complete, interoperable, and useable one. You could be a big help in
> developing that kind of spec if I could redefine the project in a way
> that you could buy into. But you would have to help me do that.
Umm... no, that sounds like a spec. for a large government system,
written by committee - a laundry list of requirements.
Now the corpus of Internet and Web specs (particularly IP, HTTP, XML) is
a spec for a PLATFORM on which one can write interoperable applications
- but as soon as you talk a spec for "a suite of integrated peer-to-peer
applications to include (but not limited to) social networking,
real-time project collaboration, content management, database
management, voting, trust/reputation metrics, complementary currency,
crowd funding, etc." - you're in the world of MS Office, or Oracle's
suite of tightly integrated applications.
About the only other way to frame things that jumps out comes from "If a
FreedomBox were used as a starting platform, the PeerPoint application
package would be added on top of the FreedomBox security stack."
Well... FreedomBox is essentially a plug computer + Debian Linux + a
collection of software with goals that sound a lot like yours. One
might ask: what are you proposing to add?
> PeerPoint is still in gestation and totally open to contributions, but
> they have to be more specific than "toss some seeds and stand back".
> Where your criticism has been specific, as in your objection to "next
> net master plan" I could see their value and I made changes.
>
>
> My sense is that PeerPoint is based on a set of premises and
> approaches
> that simply don't work anymore.
>
>
>
> So bottom line, Miles, can whatever you think is wrong with PeerPoint
> be fixed or do you think I should shit-can the whole thing and play
> some solitaire?
>
I think you should seriously rethink what you're trying to accomplish,
the approach you're taking, what you personally are bringing to the
table, and what you're looking for from others.
It might well be that you'd be more effective contributing to FreedomBox
than trying to create yet another project with what sound like similar
goals. Start with a FreedomBox and start integrating additional
applications that align with your vision. Work with a community that
already is on a similar path, rather than try to coalesce a new one.
On Sun, Jun 24, 2012 at 10:40 AM, Miles Fidelman wrote:
> design by committee doesn't work.
Design by committee may not work, but design (_and_ build, review,
adjust, adapt, discover, unfold, involving everyone during the whole
thing) by community does work and is proven to produce life-affirming
architecture, in contrast to deadening architecture produced by the
default "efficient", commercial endeavor.
A committee and a community. Both are groups of people. So is a mob,
or an army, or a corporation. What's the difference?
By the way, Christopher Alexander shows that architecture has been
essentially broken as a discipline since mid 20th-century. The
symptom? It's about everything else you can think of, except the
humanity of the actual individual people that are supposed to inhabit
(use) the work. They are abstracted away, simplified as entities that
have a set of generic needs. They do not actually exist; if you ask
who the work is for, you may know the names, but you don't know them.
So it is easy to point to current architecture and building practices
and say: this is by committee, and it does not work. This is the
"standard", commercially-sound way of doing it, and it "does work".
Except that neither actually works, anywhere, and you just don't know
it because you can't even conceive of what a beautiful world covered
with actual living space would look like. First you'd have to let the
full extent of the unbearable ugliness of what already exists sink in.
Melvin,
thanks for the link.
DOAP and the stuff it is built on like RDFS and XML are definitely applicable, but in their raw form they may exceed my competence to make use of them. I may be able to use the DOAP-a-Matic interface somehow. http://crschmidt.net/semweb/doapamatic/
Is there any way to add a machine-readable semantic interface to a Google doc, spreadsheet, wiki, or database without doing a bunch of coding?
from: | Alexandre Lachèze via librelist.com | ||
reply-to: | web...@librelist.com | ||
to: | web...@librelist.com | ||
date: | Mon, Jun 25, 2012 at 1:54 PM | ||
subject: | [webp2p] KadOH DHT experimental twitter-like application | ||
mailing list: | webp2p.librelist.com Filter messages from this mailing list | ||
|
Connection/community/involvement seems to fade with numbers - members of
a congregation vs. all members of a particular denomination, co-workers
vs. a professional community, members of a unit vs. all marines. As the
numbers grow, what might remain is a sense of solidarity, common values,
maybe a common way of looking at things. But...
I: Hyperconnectivity
How many people can any given person on Earth reach directly? Before the Urban Revolution that value had a strict upper bound in Dunbar’s Number. This number sets an functional limit on the troupe (tribe) size of Homo Sapiens Sapiens. Human units larger than this fragment and bifurcate along lines of relation and communication. One tribe grows from stability into instability, and fissions into two. In the transition to the city, humanity developed other mechanisms for communication to compensate for our lack of cognitive capacity; the birth of writing proceeds directly from the informational and connective pressure of dense communities. . .
Which is kind of my reaction to this PeerPoint thing. I know how to
review, comment, and otherwise add value to something - once there's
something to review, command and otherwise add value to - but being
asked to contribute to the initial design and concept, particularly in
survey mode (add to the requirements, add to the list of technologies to
compare and contrast), that just strikes me as a recipe for going
nowhere fast. (Might be interesting to look at FreedomBox as an example
of something with similar motivations, but a very different approach -
and ask why it seems to be getting some traction and making some progress.)
Miles
Excerpt from blog.p2pfoundation.net
Book of the Day: Promise and Perils of Highly Interconnected Systems
Book: Interop: The Promise and Perils of Highly Interconnected Systems. Byh John Palfrey and Urs Gasser. Basic Books, 2012
In Interop, technology experts John Palfrey and Urs Gasser explore the immense importance of interoperability — the standardization and integration of technology — and show how this simple principle will hold the key to our success in the coming decades and beyond.
The practice of standardization has been facilitating innovation and economic growth for centuries. The standardization of the railroad gauge revolutionized the flow of commodities, the standardization of money revolutionized debt markets and simplified trade, and the standardization of credit networks has allowed for the purchase of goods using money deposited in a bank half a world away. These advancements did not eradicate the different systems they affected; instead, each system has been transformed so that it can interoperate with systems all over the world, while still preserving local diversity.
As Palfrey and Gasser show, interoperability is a critical aspect of any successful system—and now it is more important than ever. Today we are confronted with challenges that affect us on a global scale: the financial crisis, the quest for sustainable energy, and the need to reform health care systems and improve… Continue reading »
I'm thinking the users of a piece of software are as important as the
programmers, if not more because the programmers get the programming
flow kick while the users have to eat all that initial dog food.
Quality users, users that care, are actually part of the development
effort. They _are_ developers, even designers of the original intent.
Mark, I ran across something while researching ontologies that might interest you: http://cprtheory.referata.com/wiki/Main_Page
This website is used for the project on Scaling up Common-Pool Resource (CPR) Theory. It is a semantic wiki, which is a wiki that holds data. The data in this site describe the values of variables for a set of example social-ecological systems (cases). These data are held in the form of instances of categories. The categories are shown in the sidebar to the left under "Data". For more information on these categories and adding instances of each of them, click on the "Methods" link to the left. What follows here is a brief description of the contents of this site, the outline of which is given in the sidebar to the left.Pages in category "Case"
The following 5 pages are in this category, out of 5 total.
On Sunday, June 24, 2012 1:13:51 PM UTC-5, Mark Roest wrote:
I am copying Jack Park, who may be able to offer assistance here.
On Sun, Jun 24, 2012 at 9:16 AM, Poor Richard wrote:
Melvin,
thanks for the link.
DOAP and the stuff it is built on like RDFS and XML are definitely applicable, but in their raw form they may exceed my competence to make use of them. I may be able to use the DOAP-a-Matic interface somehow. http://crschmidt.net/semweb/doapamatic/
Is there any way to add a machine-readable semantic interface to a Google doc, spreadsheet, wiki, or database without doing a bunch of coding?
The scope of FeedSync for Atom and RSS is to define the minimum extensions necessary to enable loosely-cooperating applications to use Atom and RSS feeds as the basis for item sharing – that is, the bi-directional, asynchronous synchronization of new and changed items amongst two or more cross-subscribed feeds.
Note that while much of FeedSync is currently defined in terms of Atom and RSS feeds, at its core what FeedSync strictly requires is:
- A flat collection of items to be synchronized
- A set of per-item sync metadata that is maintained at all endpoints
- A set of algorithms followed by all endpoints to create, update, merge, and conflict resolve all items