Jim Moore’s blog: Innovation, Strategy, Public Policy

Dave Winer just wrote a nice email note in response to my post on OPML at Harvard:

Jim, that’s why I went to Harvard two years ago, to bootstrap what you’re booting up now. Good work!

Why is now the time for the Really Simple knowlege Syndication revolution to take off in universities?  I have three suggestions:

1. The form of initial participation in OPML matches how academics and researchers work. 

Academics in particular continually make and share lists and outlines, e.g. annotated bibliographies, course lists, syllabus, citation indexes, research sites, research data–so I think lists are an easier starting point for participation than blogs, at least for most.

News reporters write daily posts, and naturally see blogging as valuable.  Academics make, organize, append, share, and discuss lists and outlines.  Academics quickly see the value in OPML outlines and OPML outline editors.

3.  Academics now have powerful and easy-to-use tools that enable them to compose, edit, borrow and share, and publish OPML, as well as search for it,  study it, and integrate new OPML content into existing outlines and online work environments, which in turn generate more OPML.

Fun, check this out

The future is already here.  Check out this active directory search result on OPMLsearch.com:

The search was for the term internet law, with no quotation marks, and returned the following URI:

http://70.85.87.132/opmlsearch/OPMLTree.aspx?outlineGuid=7a2705d1-e623-40cf-91e2-3f9176f35f37&Search=internet+law&Results=*&cp=0

Where did this come from? 

http://h2obeta.law.harvard.edu/home.do

It came from a particular community of use at Harvard that is developing an outline-based way to share academic knowledge, but that was a closed island of information until it began outputing OPML a few days ago. 

Now it has a connection to the world.  Or rather, now it has openned up to the OPML community, and we can connect to it’s data, suck it into our systems, play with it and share it. Bravo!

FYI, here is a link to the most popular H2O playlists, to give you an example of what people have been doing

http://h2obeta.law.harvard.edu/viewAssociatedPlaylists.do?type=influential

The work is quite interesting. It includes material from a wide range of contributors, on many topics. For example, it includes course descriptions the MIT Open Courseware project, for example.  (Unfortunately the Open Courseware project has not yet adopted OPML output, but that won’t take long once they get the message..)

Technically, adding OPML output to the site was easy to do.

The rest is history.  As soon as OPMLsearch.com started picking up the files, the knowledge based in this site became instantly available to the OPML community.  And because the work is in the OPML format, it is easily tagged, mixed, extended, shared, etc. 

In one specific case an ”H2O Playlist”–a knowledge sharing outline aka playlist–was created at a site at Harvard by John Palfrey.

Here is the H2O display of this playlist:

http://h2obeta.law.harvard.edu/62484

For comparison, the OPMLsearch active directory version is here.

What will they think of next?

Now, think of all the other communities of practice and repositories of knowledge that can be brought together in this simple manner.

The OPML is created as a byproduct of other activities and systems, in the background. 

It is so easy and inexpensive to output OPML that the development and support requirements can be justified without knowing in advance how the OPML will be used. 

And the beauty of OPML is that someone somewhere will start working right away on how to do something creative with your OPML output.

Personal reading lists..

Curriculum and reading lists, for example, from other universities..

Library card catalogues..

In a world dominated by informal media, the issue of what news sources are most legitimate is hugely important–most current, most personable, most relevant in terms of focus and selection of items, most insightful, and most trustworthy.

The world of informal media now has two main legitimizing institutions, each of which has its strengths and weaknesses.  I wonder if we might invent more?

The first are the A-list bloggers.  Dave, Dan, Joi, Glen, Markos, and Doc, etc.  The good news is that these folks hold their places in large part because they do important work for the community, and they do it diligently and well. They got where they are by dint of contribution.  The downside is that network effects have locked these folks in, and they represent a particular subset of the burgeoning world of informal media.  As has been noted in recent weeks, for example, this group is heavily dominated by men, and men of a certain age. This is not to criticize the A-list bloggers, just to say that it is difficult for new folks to break in now, and this may become a problem as the blogosphere expands.

The second is Google page rank. Non-transparent, mysterious, but very legitimizing for some purpose.  Essentially, Google turns the links that folks maintain on their web sites–the link rolls, the pointers, etc, into an implicit tagging system.  The good news of this system is that it can encompass an almost infinite range of topics, as many topics as can be meaningfully described in keyword searches.  And it is quite open in that a new, ermergent topic can be established, gather a cluster of links, and become searchable without any human intervention or permission and even notice on the part of Google.  The downside is that Google ratings do not show much personal expert judgement, they are slow to stabilize around a new topic, and tend to point to works back in time rather than current contributions. 

Note that Google News is current, but focuses on formal media and does not provide much legitimizing of sources.  Sources in the formal media world are self-legitimizing through their brands and promotion of those brands.

Of course their are other legitimizing institutions, with powerful if specialized influence.  Slashdot is wildly popular, with quickness, openness and transparency, relevance and personality in the tech space, if less effective at selection and insight.

Global Voices is playing a wonderful and unique role in pointing to and legimizing informal media around the world.

Commercial blogs are playing an important role. Think Boing Boing.  Highly pesonal and relevant and insightful. But they are generally not very transparent and open.  There is some question as to whether the advertising and editorial sides of their businesses are effectively separated.  These are commercial versions of A-list bloggers.

And tags on de.licoi.us are coming on strong, within the community of active, hard core digerati. Unfortunately these tags do not yet reach a mass audience.  BTW you know what would be cool?  An A-list blogger or team of bloggers who followed and reported on interesting developments in the world of de.licio.us and Flickr tags, and evangelized reading and writing tags, and the formation of tag communities.  Tagging for the rest of us.

In the early days of the personal computer revolution there were “evangelists” at every turn, and evangelist was a well-understood and respected role.  The most famous was Guy Kawasaki, whose title at Apple was “Chief Evangelist.”

The goal of the personal computer evangelist was simple:  Get more people in your company, neighborhood, social set, and family using personal computers.  This goal was highly achievable, as personal computers became readily available.  It was a measureable goal, the best measure being “number of users added,” and in many companies and schools large numbers of new users were added. 

At Harvard Business School, where I worked at the time, a revolution was started by providing all incoming MBA students with compatible personal computers, in 1984/85.  Soon other business schools followed.

Imho, it would be quite interesting to do the same thing with blogs at Harvard Business School in 2005/06.

Now why am I ruminating on this ancient history? Because the gating issue today for the second superpower, for citizen journalism, and for the spread of personal expression and personal syndication is still a lack of content in many arenas of knowledge, particularly outside of the US, UK, and Japan.

And for this we need more evangelists.  Dave Winer has been a model of this, of course, touring around the country in his van, spreading blogging to various centers of knowledge, including our own Harvard program.  But he can’t do it alone.

More than a year ago at BloggerCon a number of us agreed to focus on getting blogging going on, and in, countries around the world.  Britt Blaser challenged members of the group to pick a country and focus on it.  This led some of us to start Sudan: The Passion of the Present; as well, Rebecca McKinnon and Ethan Zuckerman, convenors of the session, have made Global Voices into a phenomenon.

But we need more.  We need to have dedicated, skilled evangelists pushing blogging into every knowledge space that exists.  We need evangelists to find ways to inspire and encourage new bloggers, to bring them relevant traffic, good tools, and help them develop their online communities.  And we need companies and wealthy individuals to support the evangelist corp.  This initiative is critical to the expansion of the new communication and knowledge ecosystem enabled by blogging and RSS.

It is really fun to see what folks are coming up with for using OPML, between Dave’s encouragement to put things in OPML, and the ability to search for OPML files and display them at OPMLsearch.  OPML is fast coming to mean open knowledge publishing, hosting, searching, displaying, and exploring.

Libsyn (”Liberated Syndication”) the leading podcast content hosting service has created  an OPML file of the PODcasts that they host.  Here is an OPMLsearch result you may find interesting:

Lybsyn directory displayed at OPML Search

OPMLsearch renders the directory as a nicely formated tree. Click on “+” to expand any item in the outline.

Bela Labovitch of RSS Labs posted this link this morning.

McD of the OPML community list noted this and kindly passed the link around by email.

Here is Bela’s post introducing the link in her blog at blogs.opml.org.

http://blogs.opml.org/belaLabovitch

Here is the actual URL for the search, at OPMLsearch.com

http://70.85.87.132/opmlsearch/OPMLTree.aspx?outlineGuid=2605034a-d28d-4b31-8c68-0fcdb66a1980

Digital sure beats paper..

http://www.eastandard.net/hm_news/news.php?articleid=27894

http://search.msn.com/

http://www.icerocket.com/

http://www.blogdigger.com/

Interesting question, how will these change the RSS ecosystem? What do they enable you to do that Google does not? One thing that is of particular importance to me is the easy establishment of new value chains among open web superservices.

In this light, my favorite search engine now is MSN Search.  It follows most closely the minimalist but powerful conventions of open web superservices, and enables the following value chain

spider a pool of data>search a pool of data>output search results in machine-readable RSS form>aggregate with other complementary feeds>display

to be constructed (nee, programmed) by typing one set of search commands, and clicking three times, assuming you already have an aggregator account. 

Here is how MSN Search does this:

1. User initiates an action that generates a persistent URI as well as a meaningful display of data:

The main opening page of MSN search is

http://search.msn.com/

It has a search  bar that creates persistent URIs–here is one for a search for RSS Investors (not in quotation marks)

http://search.msn.com/results.aspx?q=RSS+Investors&FORM=QBRE

2.  User can generate a persistent URI that outputs results in RSS machine readable form:

With one click on an orange RSS button, the program outputs a persistent URI for the same search with an RSS output that can be subscribed to in any news aggregator

http://search.msn.com/results.aspx?q=RSS+Investors&format=rss&FORM=RSRE

3.  The superservice makes available a one-click ‘next action’ that automatically links together web superservices into ongoing, semi-permanent syndication relationships with popular downstream superservices, thus creating a new value chain.

There is a one one-click subscription button-generated URI that takes this search and puts it into a “My MSN” aggregator

http://my.msn.com/addtomymsn.armx?m=1&id=rss&ut=http%3a%2f%2fsearch.msn.com%3a80%2fresults.aspx%3fq%3dRSS+Investors%26format%3drss%26FORM%3dR1RS

Even more amazingly, there is a Microsoft-generated subscription link to My Yahoo–a competitor’s news aggregator

http://e.my.yahoo.com/config/cstore?.opt=content&.url=http%3a//search.msn.com%3a80/results.aspx%3fq=RSS%2bInvestors%26format=rss%26FORM=R3RS

And there is one to Bloglines!

http://www.bloglines.com/sub/http://search.msn.com:80/results.aspx?q=RSS+Investors&format=rss&FORM=R4RS

This is the way services ought to work!  This is an important model for developers to study.

Trees.  Two forty-five-foot sugar maples rise from my lawn.  I look out a second-floor front window, and see at  the top of one tree a hundred or so yellowing leaves.  These turning leaves stand out within the maples’ broad summer crowns of green. 

A smaller, twenty-foot maple sticks up from a mass of vegetation behind the house.  It displays a few dozen orange leaves on one side. 

A six-foot newly planted Japanese maple by the stone steps entering my house displays a fan of stems waving orange-yellow leaves.  It’s two counterparts, planted within two yards and watered the same amount, are not turning  at all.

Three maple trees, of dozens in easy view, show a littlle yellow and orange at their tops.  Does the grand turning of the leaves, in full in mid October, always start with a subtle, gradual turning in August?

What makes these few trees the earliest?  Less water? More water?  Genes?  Network architecture of capilary systems passing nutrients to top-most branches?

We have a hint of possible rain today, with high, fast-moving scattered clouds.  The yard brightens and dims and brightens again every ten minutes or so. 

Just published yesterday in Forbes:  Tom Taully has a comprehensive piece on “The Rise of RSS” in Forbes online, August 17, 2005.  Excerpt:

NEW YORK - What’s the hottest brand on the Web these days? The orange RSS icon.

Take a look at the online sites of any major media company. You will see hundreds of these icons offering RSS, or Really Simple Syndication, which is a way to get information sent to you, such as articles, classified ads (say from Craigslist.com), hotel schedules or even audio files (or podcasts).

He interviewed me about the potential for RSS to transform the enterprise, which I am enthusiastic about.  Thanks Tom!

Tom illustrates that something big is happening, using a well-selected roundup of initiatives (such as Microsoft  support  for RSS), deals (Bloglines and MySpace),  companies (Rojo), people (Andreessen, Conway) and VC firms (Sequoia and Kleiner, Perkins).

DIY citizenship

There is a new, global DIY do-it-yourself revolution happening. 
My new favorite magazine is Make, http://makezine.com which is a
wonderfully screwball journal of hacks–e.g. “open source cars”–that
is totally fun and inspiring. In a way, Make summarizes for me what is
going on around the world as the number of educated, creative people
expands dramatically–reaching into at least the hundreds of millions
of souls.

DIY Superservices

In information and communication technology, DIY do-it-yourself
revolution is expressing itself, and enabling itself, through
do-it-yourself open superservices made available on the web–and a
massive, parallel expansion in the number of individuals who are using
these superservices to create new applications, many of which are in
turn made available as new superservices. These are “DIY services” or better,  “DIY Superservices.”

These services are super in many ways, including:

Web superservices are capable in theory–and sometimes in
practice–of scaling to serve large numbers of users and transactions.

Web superservices can be created by anyone with access to the web. 

Web superservices are open to all users across the planet. 
Superservices are user scriptable in most instances.  Users can
combine services into networks of intwined inputs and outputs, creating
service clusters that are far more valuable than the sum of their
parts.  User-created hacks and mash-ups are becoming commonplace.

Web superservices reflect the imaginations of their developers, and
are not forced into any typology, registray, or typing system.

DIY Superservices are not traditional corporate web services

Thus while superservices exist on the web, and they are services, and they take
inspiration from traditional web services, superservices have sprung
free and are fast evolving into something very different from their
ancestors.

Traditional web services were tightly typed and constrained by the
typology within which
they were developed.  Traditional web services are of necessity
sponsored by large corporation because only a powerful economic entity
can fund the design of a grand service typology, enforce a complex
implementation of typing, and establish the training and tools and
conferences and other institutions needed to establish a community of
developers willing to work together under a particular framework. 
This is why there is really no competition in web services to the two
corporate enties, Java, supported by Sun, and .net from Microsoft.

The new world of web superservices has no
such limitations.  Web superservices are by design only minimally
or loosely typed, in order to promote the freedom of developers to
imagine what they will. This is the big idea.  The less typing the
better. The less change in the basic standards–for example, URLs and
RSS–the better.  Loose typing promotes community participation.

For those  who are “Snow Crash” fans, the new world of
superservices is more akin to the Metaverse, Stephenson’s imagined
digital city that has no boundaries, and allows every hacker to build
whatever house or building he or she desires.  The city expands
without limit as new structures are built on open ground at its edges,
and the
edges move out to encompass more territory. 

Infrastrurcture in the metaverse springs up like a-list bloggers in the bloggosphere.  Certain nightclubs
become central gathering places.  Roads and corridors lace
through the Metaverse, traversed by motercycles that could travel at
nearly infinite speeds.

Fifteen reasons why DYI Web Superservices will transform the landscape

Many people observe that we are in the fastest information and
communication technology market yet seen.  Why is this so? 
Here, for your consideration, are some things I have observed about the
virtual world that is emerging around us:

1.  We are planting and cultivating a new open ecology of web superservices.  Or perhaps we are joining together
to build a Metaverse in
Stephenson’s sense. In any case, expanding participation is rule number
one of our new world.  Open ecologies are inherently complicated,
co-evolving
combinations of counter-intuitive relationships. Thus wild open
ecologies, when they have the conditions to take off, always outrace
more structured forms of cultivation.  Do you have to plant
weeds?  What happens when weeds meet your garden?  Which
ecological form of organization wins if you don’t intervene?

See del.icio.us as an open planetary web superservice.

http://del.icio.us/

These new services are novel, fun, and often solve important problems we didn’t even know we had.

2.  These web superservices are enabling new forms of community
cooperation by enabling vast shared pools of knowledge contributed to
by dozens, hundreds, thousands and sometimes millions of users. Of
course the web itself is the largest of these pools, but the
interesting development is the creation of more and more specialized
pools, still having a critical mass of contributors and a valuable
shared resource of information.

Read about ajax on del.icio.us, for example, to explore a small but potentially interesting pool.

http://del.icio.us/tag/ajax

3.  We are experiencing the return
of the command line in computing.  The URL has become a the
command line for open superservices.

The classic Google interface, for example, is now seen by web
superservices hackers as a command line generator.  The Google
interface is code generator.

Here is an example:

http://www.google.com/search?hl=en&q=geocode&btnG=Google+Search

This command causes the Google machinery to perform a search and render the results in a particular manner.

4.  “Pull” relationships among services have now become powerfully
generalized.  We live in a pull universe.  Open pull has
become the default way to design access to a new web
superservice.  The simplest and most powerful version of this is
the RSS file, where an application such as blog content manager creates
a version of the blog content as any one of a number of XML dialects,
and then deposits that file in a folder that is open to the whole web,
and thus can be reached by a URL “command line” and pulled out for use
by anyone. 

See my file for this blog.

http://blogs.law.harvard.edu/jim/xml/rss.xml

5.  Pull enables community creativity.  There are millions of such RSS sources on the web, untold numbers of
other open web superservices that can be reached by a URL and provide
an output drawn on by anyone anywhere anytime.  This is a very big
deal in terms of enabling creativity.

6.  Pull equals more web superservices introduced.  A creative service developer needs very little
to introduce a service.  All it takes to introduce a service is an
application that can be evoked by a command embedded in a URL, and that
can provide an RSS or other output that is machine readable and
deposited in a file open to the web.  Presto, you are now the
inventor and publisher/provider of a new web superservice!

7.  Pull equals more users of superservices.  A creative service user needs
very little to access a service. She needs a URL, or perhaps access to
a web page that generates URLS, and he needs to be able to read the
output.  Both functions are built into every browser.

8.  Pull enables the creation of complementary tools and
services that in turn expand the range of inputs and outputs to core
services.  Command-line generators, scripts expand teh application
interface of superservices.  Output readers, parsers, format
transformers provide connectors to traditional applications and to new
complementary web services, all expanding the ecology around any given
superservice.  Thus the podcast ecosystem now connects podcasting
production sotware, search engines, and podcast downloading software
that link to MP3 players and iPods.  The community of creative
service users is  free to innovate in terms of how the put
services together into end-to-end solutions, and free to build
components to enhance these solutions.

9.  Pull enables scripting together services, and thus more
superservice-based integrated applications.  A network of
pull-based services can be scripted together
simply by a set of users sharing URLs, as I have done in my notes on
this page,  and other users plugging those URLs into their
browsers, reading the results, and passing them on by way of blogging
or some other form of downstream publishing.  Moreover, automated
networks of web superservices can be easily scripted with using PHP or
other simple languages built for this purpose.

10.  Pull reduces spam, spoofing, and other authentication problems. 
The user of a pull-based service must elect to reach out to that
service, thus such services are inherently “opt-in.” Because the user
knows the source address of the service, the service is
self-identifying and (to some extent) self-authenticating.  The
pull-based nature of web superservices is the opposite of email, where
a push-based architecture allows spam and anonymous sources to be a
huge problem. In the push world of email the user maintains an open
folder to the web, and any source can push junk to the user.  In
the pull world of web superservices that source maintains the open
file, and the user must reach into that file.

11.   No namepace need be maintained by publisher.  We can see the brilliance and elegance of a pull-based
universe of superservices if we take note of what is NOT required under
approach.  The elimination of each of the following “nots” reduces
barriers to creativity and feeds the proliferation of new applications
and the collective expansion of the user community. In a pull-based web superservice the publisher does not need to
establish and maintain a namespace and a list of permitted users. 
The traditional telecom service function of provisioning  goes
away.

12.  No centralized publisher registration and/or strong
typing need be conformed to.  The publisher does not need to register with anyone or
anything. There
is no master framework of design guidlines.  There is no service
typing in the conventional sense.  None is needed.  With a
URL for input and RSS for output, powerful services can be
created.  Of  course other forms of input and output are
allowed, for there are no restrictions.  If users like these new
forms of output, the network of services will be extended.  For
example, geocoding is making available map-based outputs. Obviously
these outputs are gaining traction in the user community, and we can
expect to see services start to proliferate that will have geo-related
inputs and outputs.

13.  No user registration is required.  The user need only have a URL and
a way to pull a file from a location on the web, to get access to a
service.

14. Downstream transformation is encouraged.  The new world provides vast user freedom to integrate, transform,
and republish content.  Once the user has hold of the content,
there is no technical or practical barrier to doing what one will.

Of course there are thicket of unresolved legal issues related to this
freedom.  And DRM systems are evolving to try to limit this
freedom.  Copyright licensing superservice Creative Commons makes
it easy for publishers to give permission for a wide and controlled
range of downsteam uses.   And technology continues to enable
breaking DRN locks, for better or worse depending on your point of view.

As a practical matter, there is almost no limit on how fast and
how far the community can go with the integration, transformation, and
republication of the outputs of web superservices. Mash ups, sampling, repurposing of content are inherent in the new world.

15.  Syndication is a core idea, and perhaps the core idea.  The new
world establishes in the most powerful and general sense the idea of
syndication.  Syndication describes a world of
user/publishers.  Every user can subscribe to the content produced
by other user/publishers.  The user then integrates and transforms
that content and publishes his or her own content.  This in turn
is picked up by others.  This great process of syndication is
essentially what happens in language, in face-to-face communities, as
people pass on ideas to each other, transform the ideas and in turn
pass them on again.

What is powerful today is that we now have a world of tools that enable
networks of user/publishers to process immense amounts and diversity of
content.

Now the tools themselves are evolving into superservices that can
syndicate each other’s content, at the command of users.  Thus not
only can users become user/publishers–the blog and aggregator
world–but users can now become scripters of clusters and networks of
automated and semi-automated user/publishing services.

—————————————————————————
Link to this article: http://blogs.law.harvard.edu/jim/2005/08/17#a1007

Subscribe to this feed:  http://blogs.law.harvard.edu/jim/xml/rss.xml

“>