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

We each have an interest in the knowledge in the professions. Medicine, in particular, touches each of us and every family member. Knowlege management in medicine is a matter of life and death.

The outline is central to providing access to knowledge. The annotated bibliography, the reading list, the journal citations, the directory of research literature, the citation indexes, and so on are core structures for the organization, sharing, teaching and learning, and use of knowledge in professions.

I believe that organizational learning has been vastly enriched by access to online sources.

On the other hand, the abundance of materials available to us today cry out for organization and structure–something more constructive–or better said, more constructed–than lists of search engine results.

One answer is found in the creation of digital directories–the online, digital versions of indexes and annotated bibliographies.

The special opportunity with OPML is that it is a computer language created for outlines. Thus as the OPML tool set is expanded, we are making a world where outlines can be more and more easily used. And the hope, at a higher level, is that this in turn makes knowledge more easily used.

So here is a profoundly interesting and useful OPML-based reference application created by Philippe Campeau and posted tonight on The Online Metabolic and Molecular Bases of Disease Blog.

The OMMBID Blog » Blog Archive » Journal articles
Created using Internet Explorer 7, Google reader, Intelligentteams.com and Grazr. Use the buttons on top of the reader to change the view or open in a new window. Have updates of these feeds emailed to you using RMail.

Philippe Campeau

Wrap your mind around this: The collective expertise organized in this application is immense. Yet today’s tools, using OPML, make it a relatively simple matter to make an active, continuously-current, and automatically-updated widget that provides near instant access to this collection of materials. This widget, in clone, can be easily placed on any blog, start-page, web-page (say, Google Pages) or web site. And wherever it lives, it will stay in touch and current with its OPML source.

And all this can be done by the professional him or herself.

Thanks Philippe!

Today, a chapter on Video!

Intelligent Teams/OPML Workstation is an OPML site that is optimized for embedding short video and other widget-based media.

The future of the web is clearly in short video. Video is the new web design medium.

The future of OPML = the carrier for video. OPML is the open, public format for video websites and for videocasting.

For example, here is a simple but effective combination of OPML and video in a small video website.

Consider the following:

I. Why is video the future?

1. Video is catchy, viral, and conveys emotion more effectively than all but the most well-crafted text. Small videos are viewable on cell phones, iPods and other media players, and PDAs–as well, of course, as on PCs.

2. Video is accessible. Users can view videos more easily than they can read. Nuff said.

3. Video is easy to make. I make mine on an old Cannon Elph still camera, using its “movie” feature. New still cameras make mpg4 videos, of up to two hours in length, on one 2gig flash card. Panasonic and Sharp both do HDTV-like aspect ratios, look-and-feel.

4. Video hosting and serving/bandwidth is free at YouTube and a bunch of other sites.

5. The “video layer” of the web is developing fast. Most of the buzz is now about videos. Not audio, and certainly not about text.

II. Why do we need an XML-based carrier for video?

Discrete videos strewn across the web are more like raw material than like actual destinations or programmed experiences. The same was true of mp3 audio files prior to the invention of podcasting. Podcasting provided context–lists, directories, subscriptions. Podcasting provide the ability of individuals to make their own customer context–playlists, personal directories, and subscription lists. Podcasting provided ways for users to share these lists.

III. What happened to open audio, open music, and open podcasts?

Audio does not have an established open XML-based carrier standard.

Community innnovation has been thwarted in the audio and thus the podcasting world by Apple. Apple adopted a non-standard, closed format for the XML meta-data layer that organizes its library of songs and podcasts and feeds its menu systems. Apple could have used OPML or some other open standard, but did not.

Apple’s underlying music files are in a proprietary format, they are covered by a highly restrictive DRM (read today’s NYT essay on this issue), and the files themselves can only be accessed by way of a downloaded special software client. The files cannot be accessed by a browser on the open web.

It is thus very difficult to create independent web sites that incorporate Apple-controlled content. If Apple made music and podcasts available openly on the web, things would be different.

The good news, as we will explore in more detail later, is that user-generated video is already evolving in a different, more open manner. As a community we should jump on this opportunity. Part of this is due to open design choices made by the video hosting sites. YouTube, Google Video and other popular video sites make their files available for direct, unmediated web access, with no DRM.

By contrast, a closed trend is solidifying as downloaded music, podcasting and mobile phones converge. Phone companies seem intent on adopting closed systems for cataloguing, conveying/distributing, and selecting songs.

IV. Why did the blogroll, RSS subscription list, reading list ecosystem evolve in a more open manner?

Community innovation in the parallel world of text-oriented RSS “reading lists” has been much better supported, because vendors have more or less standardized on OPML as a way to share sources and lists among

1. Blog aggregators like Bloglines, Newsgator, and Attensa;

2. Feed proxy services like Feedburner;

3. OPML authoring and hosting services including web-based Intelligent Teams, and client-based authoring by Dave Winer’s OPML Editor and Pito Sala’s Blogbridge;

4. Search services like OPML Search;

5. Source-popularity and sharing services like Bloglines’ Most Popular Feeds and Bloglines Share, and Share Your OPML;

6. List-based dynamic-content analysis services like Megite and TopTenSources;

7. Social bookmark services like Stylefeeder that create OPML lists (TopTenSources provides the most powerful bookmarking and sampling technology I have seen, including one-click aquisition of YouTube videos, in a free tool you can get by signing up to create a TopTenSources topic or page.);

8. Community production, editorial, directoy and membership sites such as Intelligent Teams, TopTenSources and Lisa William’s Placeblogger, as well as more focused yet massive directories created at such sites, such as the by-now-famous James Corbett Open Irish Directory and the cryptic but fascinating “pro” sites of Biotic;

9. Feedlist display sites such as the AJAX-rich Grazr, Intelligent Teams‘ “browse” services and widget maker.

10. Outline-based blogs, either using OPML backends or adopting OPML-like outline conventions. See an example of an “almost outline blog” here, where a reading list, by the addition of commentary and other forms of references, is fast evolving into a powerful, directory-enhanced, visually-appealing site. A thoughtful discussion of this issue, at this point in time, is available here by John T. of Library Clips.
11. Outline-based web sites and public directories.

12. OPML power-users and pioneers, such as Harvard’s John Palfrey and Library Clips’ John Tropea (also mentioned in the next section, because John is the author of this well-regarded commentary.).

13. OPML industry commentary services such as Library Clips,

and enterprise 2.0 commenters like Charlie Wood who get OPML.

The good news is that in the case of user-created text, OPML–an open, stable, XML-based standard–is succeeding at enabling an ecosystem of creative people, products and companies to work more or less together, to interoperate, to exchange, and to co-evolve.

The bad news is that the world of text and reading lists is pretty small. To say it perhaps too bluntly, only a small percentage of web users prefer to read. Probably no more than a few tens of thousands of participants are engaged in reading RSS sources to the extent that they value a reading list. Compare this to the millions of participants in the iPod/iTunes universe who value and regularly use audio playlists.

V. How does video provide a new field of action?

Video provides a new opportunity to combine openness with richness, reach, beauty and immediate emotional impact

Videocasting already is huge, supported by channels and subscription services on YouTube.

Open, XML and OPML-guided videocasting can also be huge. Its advantages are that the creator of a videocast can be in control of the context in which his or her videos are seen. YouTube is well-engineered, but almost any web designer can make a better surround for a particular set of videos and a particular audience. Videocasting, using OPML, can provide a way for designers to create rich surrounds and contexts. They can do so quickly, as prototypes or as permanent installations.
At this historical moment, the world of video is quite open to the development of a community-based third-party, independent directories.

The content on YouTube and other video sites is diverse. The videos available range from investment analysis and business presentations to the classic, cat-on-a-toilet offerings. Budding acting troops have serialized shows. Rap and rock videos abound. YouTube is about as open as the web. Contrast this to iTunes, where formal catagories and commercial offerings dominate.

VI. A simple call-to-arms

Take advantage of the simple-yet-power OPML tools available today. Create contexts for videos.

Video viewing lists. Start by making lists and blogrolls of favorite videos on YouTube or similar site. This method is fast, simple, and provides some degree of context.

Multi-media-based OPML web sites. But once you are going, you can do so much more! Use OPML frames to combine forms of media–images, audio and video–on sites you can create in an evening.

For fun check out http://www.davemoore.info. I made a video site for my brother Dave while recovering in Iowa from back surgery (”no heavy lifting, all indoor work”). The original site is at http://intelligentteams.com/browse/aboutdavemoore. Then at Dave’s suggestion we bought the domain name for “davemoore.info” for $1.99. Best deal of the year.

So visit Dave! (Family summary: Dave and I and our third brother Charles grew up in Iowa. Dave went south, and then further south, and eventually planted his deep roots in Iowa City, from which he continues to range widely. Charles ranged widely as a quarter-horse trainer and outdoorsman, settling in Wyoming, by way of most of the states of the American southwest. Our mom passed away in ‘94. Our dad is lively and creative and living in Iowa.)

VII. How can we seize the moment?

1. Put anything you like up on one of the video hosting sites such as YouTube. Put up political video, home and family, professional, music and entertainment, comedy, technology, space travel–you name it. The new video platforms exist for you!

2. Create distinct, personal web micro sites that reference those videos and other offerings.

Here is a brief “how-to” to help you create rich web pages in OPML:

a. Open an OPML outline in one of the many OPML writer/host suites, such as Intelligent Teams.

b. This outline, when displayed in the “browse” function on your OPML host site such as Intelligent Teams, or when displayed in an independent AJAX OPML viewer like Grazr, becomes an outline-oriented web site. Each node in the OPML file is now a page on your new site.

c. Each of these OPML nodes can now be treated as a distinct page that can function as a mini-site. Each page can be developed using plaintext, HTML, and scripts–if your OPML writer allows them (see item f, below).

d. How do you do this? You simply write your desired web page code into the “text” field in any of the OPML nodes. This works because when OPML files are read by OPML software, the “text” field is not parsed as XML, but rather is preserved whole and passed through to the user’s Firefox, Explorer or other webbrowser to be interpreted and displayed at the time it is loaded.

Thus material written into the “text” field does not need to conform to XML rules. Anything that the final, end-user browser can interpret can be included effectively. This means that each node’s “text” field can be used as a defacto iframe-like capability, into which one can encode rich web page material.

e. To include YouTube or Google video in your pages, simply copy the “embed code” provided by the hosting site into the “text” field of a node. The result is that a video player–typically Flash-Shockwave, for example–and an associated video are coded into the outline page and displayed when that page is subsequently viewed by an audience member/visitor to the page.

f. Use script-friendly OPML tools. In order for this use of OPML to be effective, your OPML file must viewed in an OPML displayer that allows scripts to be executed by the end-user browser.

Some OPML displayers and many blogging software services remove scripts. For example, the WordPress version on which this post is being constructed blocks scripts, which is why there is no video in this post. Intelligent Teams is an example of an environment that allows scripts in its writer as well as in its viewer. Thus it can handle not only video, but most widgets, as well.

3. Add display customization and personalization including text and skins that display those videos in a context of your design. Now you are free! Your materials are no longer enslaved to the look and feel of YouTube, flicker, or any other site. Now your materials enhance your minisite and your brand and image, not simply that of the file hosting platform.

4. Because you have used OPML you now have a portable web site design, viewable in Grazr, OPML Workstation, or any one of a host of other viewers.

When you use OPML you gain portability and transformability. Indeed, you will be able to view your design simultaneously, without altering its core code, in widgets, gadgets, mini-sites, aggregators and so on. No need to choose just one. See John Palfrey on OPML for Teachers, in a widget-maker, for example.

VIII. What is the key transformation we are making in web site architecture?

We are making web sites that take as a given, and as raw material, a rich understory of graphics, text, music and audio, and especially video. Web sites reference this material, organize it, shape its presentation, and guide the experience of each audience member.

From a technical standpoint, the new web is characterised by write once in a display and directory language like OPML, reference many (videos, audios, data sources), and display broadly and prolifically, everywhere.

Write once, reference many, display everywhere!

About Dave Moore (songwriter, roots music)
I am back to writing a new OPML chapter a day, after a two week pause. I have a good excuse. I had two (yes, two) emergency back surgeries.

The first surgery was on Friday, December 22, at the Mayo Clinic in Rochester, Minnesota.

On Saturday just before Christmas my dad drove up to Rochester. He kindly brought me back to eastern Iowa, where I grew up and he still lives.

We arrived on Christmas eve. For Christmas eve dinner we were guests of his generous next door neighbors, Tim and Sara Kelley, and their son Sean, daughter-in-law Rachel, and two grandchildren Brian and Reilly. A very creative family.

Rachel Williams is a professor of Art Education at the University of Iowa, and she and I got to discussing online portfolios. OPML is perfect for online portfolios–so I was excited by her ideas. I have long been a fan of portfolios as an approach to summarizing the accomplishments in education. Portfolios are much better than grades. With portfolios, students are encouraged to make real, transferable objects, true and lasting contributions, and present those to prospective collaborators, customers, and employers.

The next day I seized an opportunity to experiment with an OPML portfolio. My brother Dave is a roots singer-songwriter.*

Dave has a variety of music that is available online, including samples of songs on Amazon.com, downloadable songs on iTunes, and videos on YouTube and Google video. The issue is how to put references together for these materials, and how to make a site that is friendly to the user that provides attractive and easy access.

Christmas day my Dad and I journeyed to Iowa City, for brunch with my brother Dave, sister-in-law Lysa, and daughter Josie. We enjoy a grand spread, good fun, and a long walk in the woods along the Iowa river, following Dave’s two dogs who romped ahead.

Somewhere in the walk I suggested Dave and I do an interview and perhaps make a portfolio of his work. Since I was laid up anyway, we waited a couple of days and then met down at my dad’s house. The interview, done with an old Cannon Elph, SD110, 3.2 megapixels, turned out great.

I combined this with some other material, in a portfolio created on intelligentteams.com.

Click http://intelligentteams.com/browse/aboutdavemoore. Enjoy! This is a work-in-progress, and I expect to add more material as I have opportunity. That, of course, is the nature of a portfolio.

*We three brothers are close personally but very different professionally. Our other brother Charlie became a professional horse trainer and now is a businessman and blogger in Casper, Wyoming.

OPML Search shows a fun example here of open playlists–do you like Blues music?

The Playlist Layer of the Web

The structure of the web ecosystem is changing rapidly, with a new layer emerging on the web: the playlist layer. At this point the playlists that make it up are fragmented and mostly proprietary, but a world of open playlists is as inevitable as the commoditization of any other layer of the web.

In the next world, meaning now, folks do not want “search results,” they want playlists. They want these playlists to be easy-to-use, dynamically-updated, relevant, of high quality, and easily accessible. They want these playlists to be open to innovators–that is, to content producers who may not work with mainstream media companies. On the other hand, they appreciate “curating” and list editing, reputation services, and in general, human intelligence (humint) applied to sorting out wheat from chaff.

The iPod is not a music player. It is a playlist manipulator. It is the most elegant playlist machine available. Tivo is not a video recorder, it is a video playlist machine.

The competitive evolution of Playlist Ecosystems

The ecological elements necessary for a successful playlist experience include much more than the playlist. The experience includes search, “playing” and display, etc. Indeed, it includes much more than those even. The experience system, the value chain, begins far upstream, with creative communities. And it moves far downstream, or rather, it moves closer and closer to the user/experiencer–it literally moves into the earbud or video display.

Each of the major playlist leaders is seeking to maximize the effectiveness of its playlist ecosystem. As we will see below, many elements are necessary.

Each leader must find one or more dimension of the ecosystem to excel on, in order to gain customer interest. Each leader must make sure that its ecosystem is not seriously deficient in some aspect of the ecosystem, such that it cannot meet basic customer needs.

Each ecosystem must achieve operational integration to function reliably and efficiently.

The ecosytem must achieve “experience integration” so that the customer can be as passive as possible and yet gain the experience he or she wants–including, ironically, interactivity. Does the person with the iPod want to groove on the music, or play with the controls? Note, by the way, the success of the iPod shuffle. No controls, all groove.

All of this effort comes together at the customer experience–and indeed, the customer experience has multiple dimensions, including pre-purchase, use, and use-justification and sharing..but that is another story yet again.

For now, consider that each of the major playlist competitors is seeking to maximize the “wow factor” in its version of the sytem. The wow includes the “intimacy factor.” A furniture maker recently said to me that the thing that architects don’t understand about furniture is that a piece of furniture defines a world of intimate textures, of touch and smell and color, of feel and heft and a range of postures..furniture is intimate.

Apple playlists are very high on intimacy. As are all high fashion accessories. Indeed, the bag, jewelry, shoes and other accessories business–say, Gucci–is about the creation of small, especially delightful worlds of intimate experience.

Core elements of a Playlist Ecosystem

Creative community

Content design

Content production

Content distribution

Device design

Device production

Device distribution

Storage

Addressing

Source targeting

Search

Playlist display

Playlist saving/tagging

Playlist sharing

Play selection

Play

Experience

Grooving

Community grooving

————————————————————-

Ecosystem-to-ecosystem competition

The most important competition in the world is the fight for playlist ecosystem leadership. This competiton is driving billions of dollars of investment, including research, product development, and acquisitions. Why is this fight so important to the competitors? Because it is a fight for control of access to digital culture. That is, access to digital knowledge, digital music and video, digital commerce.

Companies are the strategic movers and investors, but the unit of competition is not at the company level. The companies in turn are seeking to create winning ecosystem designs, ecosystem installed-bases and incumbancies, and ecosystem-supporting-user-communities. The companies are competing, ironically, to collaborate. To win they must collaboratively co-evolve the most powerful and inclusive collection of content, partners, users, and experiences.

Consider some of the major playlist ecosystem definers in this light:

Google, Yahoo, Microsoft are competing with each other to make playlist ecosystems based on their strengths in search results

Their ecosystems produce dynamic playlists that are profoundly inclusive and comprehensive. This is a strength and causes them–or at least Google–to be the first stop on the web for those seeking playlists. It is also a weakness, because results are very broad and of uneven quality.

When you think of a playlist with a million elements, you realise why the end of search-results-as-we-know-them is just a few months away. This problem is an opportunity for Rollyo and other playlist-from-search-results companies.

Google and Yahoo ads are perhaps the most successful of playlist ecosystems. They are richly supported by the “user-generated content” of advertisers large and small.

Essentially, text display ads are small e-commerce playlists, targeted nicely, convenient display, with quality discipline by the advertiser pays business model and the requirement that ads be clicked-through to continue to be shown.

iTunes playlists and the music industry

A playlist ecosystem that found a way to graft onto and extend one of the most important pre-web playlist ecosystems–that of the recording industry. Pay-for-play, payola, “heavy rotation,” “album-oriented-rock”–all playlist inventions of the recording industry working with its marketing arm, radio.

ITunes is a proprietary format, but very open to content producers, including not only the record industry but bloggers. Very good wow factor, intimacy and identity factor at the experience-end. Steve Jobs knows how to combine wow and intimacy.

Mobile phones with music download and streaming playlists

This ecosystem is the big sleeper in the competition. It is grafting playlists onto communication. A very powerful incumbancy. Proprietary formats, proprietary display, inconvenient and comparatively closed to content producers–but with many many many more devices in place than any other platform: an up-and-comer, obviously. iPod has about 90% share of its narrowly-defined class of devices, but it has a very small share of the digital player market–and it has no streaming capability. Phone companies don’t get “wow” very well, but Motorola and Nokia are teaching them.

Steps Toward an Open Playlist Layer

So what does all this have to do with XML and especially OPML?

Create a seamless, interoperating world of playlists and their embedded perspectives

First, we need to be able navigate within a seamless realm of playlists, in order to be able to find content of interest to us.

OPML is a language of playlists, expressed in XML. We need to be able to translate playlists out of their proprietary formats, and into shared formats that allow for mixing, matching, comparing, extending, analyzing and improving playlists. This needs to be done in machine formats, not by having web users read playlists in one format, say iTunes, and type information into another format.

XML was originally designed to connect two or more databases where the formats were incompatible. XML has become much more. It has become a universal language for expressing web content, a language that enables transformation of web content, a way of syndicating across networks (e.g. RSS), and a way of carrying references to content stored in open repositories with permanent addresses (e.g. RSS enclosures and podcast audio and video files, as well as OPML attribute-based references to URLS). XML has always been conceived as dynamic, that is, as constantly changing based on whatever content it expresses or links to. Thus XML is purpose-built to be the core of a dymanic meta-layer above open content on the web.

Second, we need to make plain the perspectives embedded in any given playlist. We need to reveal the assumptions implicit in and expressed by a playlist. By contrast, today the most widely-used search engines of the static web and the “meme-trackers” of the dynamic blog world proudly hide their algorythms.

Thus today’s search engines have at best a hidden bias in their results and at worst hidden censorship. When the logic of a search engine cannot be examined, and/or when there is no practical way to compare search engine logics with alternative approaches, authors, and purposes, we are at the mercy of unseen and unspecified/implicit world-constructing-logics. These logics shape our perceptions of the online world, and they do so at the very moment we are least reflective–at the moment that we truly need to know something.

By contrast, OPML playlists can be open, they can have identified authors, and if they are machine-generated their algorythms and rules can be published and examined. The most efficient way to accomplish the meta-analysis of playlists is if they are in the same or similar formats, and XML is the obvious and probably the only reasonable format to choose.

Open up silos of exclusive content

Content elements such as audio and video files, data elements, and digitized texts need to be accessible freely and directly from playlists. Our vision should be that all content elements are reference-able from any playlist in any format operating from any place on the web. We need to clear the fog from the basic stuff of the web.

Currently, most content is isolated into silos. Blog content is an instructive exception, showing us some of what is needed, and proving that an accessible content layer can be achieved.

Blogs content–”posts” are easily reference-able because they have “permalinks” on each post, and these permalinks, as well as the URLs for the HTML pages and the RSS feeds from the site, are open to anyone on the web.

The simplest form of playlist in the blog world is the hyperlink citation from one blog post to another. Indeed, a playlist can easily be developed and maintained with href lists in blog posts.

Aggregators such as Bloglines are playlist management services for users who read blogs.

The web itself is architected to be open in terms of content. Universal access to elements of content, with a universal resource locator system–yes, I’m smiling, this is what a URL is–is the nature of the web. But over the years the open nature of the web has been reduced as search engines and other tools have become necessary but black-box intermediaries between content and user.

Harvard and the Golden Disk

Today the worst silos, from an open culture standpoint, are those where data is exclusively bonded to curation and search. For example, Harvard is digitizing its entire library resources. Currently Harvard is paying the cost of digital cataloguing, physical preparation and shipping of books to a central scanning facility. Google is paying for the actual scanning. For this, Google is returning to Harvard a “golden disk” of digitized material–but Harvard is precluded from making these contents available outside the university unless pays Google for each page thus distributed.

Meanwhile Google is keeping its own copy of the golden disk, and will make available the same pages–but only if they are accessed through its search engines and whatever business model it deems appropriate for itself. Thus the contents of the Harvard libraries will be digitized and stored, but not open to the playlists of the world’s scholars unless those playlists are created within the Google playlist ecosystem.

One would obviously hope that the contents of the second largest (after the Library of Congress) and most global library in the world would be made open to playlists of all sorts.

Indeed, the “right” answer is for Harvard to set up the contents on open servers, with permalinks as in the blogging community, and let anyone with a blog anywhere make their own references to the documents. These blogs in turn would provide the first universal playlists to the world’s richest trove of written knowledge and wisdom. Bloggers could share references and lists of references, and the result would be a universal library, curated by an unlimited number of curators, working from anywhere in the world.

Public Outlines authored by jcorbett, click here to enter a vast and endlessly fascinating ecology of playlist-of playlists providing dymanic, continuously fresh engagement with Irish personalities of the online world.

The base of the web is made up of millions of URLs. Each URL calls forth resources from web servers. In some of the most interesting cases, these URLs go to MP3, video, podcast, and 3d files. Millions of individuals tend to these gardens, making them as varied and delightful and personal as possible.
Search engines search for these files, catalogue them, and make them available to the rest of us.

But since the early days of the web there has been another way to find resources: Directories. Indeed, before search engines there were directories. Think Yahoo. Many directories have persisted and grown, despite their lack of visibility.

Lately directories have made a comeback. Podcasts, podcast directories, reading lists, Apple iTunes playlists are all directories. Why are directories helpful? Frankly, they can be much more targeted and helpful than search engines, their structure is transparent, and complex sets of results can be ordered and made available in a practical manner. You can with relative ease traverse a directory tree down three or four or more levels. And we are now seeing cool directory-enhancing software like Grazr, which is specifically intended to facilitate “grazing” across levels of deep and rich trees. Trees are coming back.

OPML is the open language of trees. OPML is the Open Playlist Markup Language of the web. OPML is the leading edge of an XML layer that is growing inexorably across the web, on top of and as an alternative to conventional search engines. Indeed, most of the innovation in search now is putting search results into trees–Rollyo–as well as using trees to limit and focus searches. Hybridization of trees and search.

Finally, the rise of “direct navigation” promoted by companies like Name Media is a direct assault on search engines by companies that provide branded directories, such as Photography.com. As these sites have proliferated users have responded, with increasing frequency typing search terms directly into the URL bar, and typically hitting a directory run by one of the direct navigation companies.

The web is changing once again. A new layer is developing. That layer is all about directories as a primary structure for organization access to web resources. It is about separating resource objects from structures. It is about having multiple structures pointing to the same resources. It is about the web as a layer of objects, “described” by outlines. It is about venn diagrams as the new way to think about the web. Venn diagrams that surround sets of resources. Venn diagrams that overlap each other. Venn diagrams that themselves can be searched and sorted and made sense of. And most important, Venn diagrams that are the results of user creativity, collective user creativity, of millions of people.

At the top of this post I highlight the Open Irish Directory, a famous and important set of sets written by James Corbett. Enjoy!

Internet Archive
The Internet Archive is building a digital library of Internet sites and other cultural artifacts in digital form. Like a paper library, we provide free access to researchers, historians, scholars, and the general public.

Knowledge is the new free.

The world economy organizes itself around free resources. Yesterday morning I had a long breakfast conversation with Doc Searls. In addition to being a font of free knowledge himself, Doc tantalized me with an image of the golden age of construction, when timber was free, and stick houses across America were built from it. Now we are in the golden age of knowledge, and each citizen with an iPod carries in her pocket a personal mansion of knowledge, stick-built from digital files.

Knowledge is the new free. What does this mean? It means that the timber of intelligent action is abundant and at hand. The challenge for us is not so much “how to know,” as “what to do” with our knowledge. What design considerations will we apply to our houses of knowledge? Will our stick-built mansions of knowledge be sustainable? Will they harmonize with the planet? Will they contribute to peace and mutual understanding across the vast diversity of human values and life styles? Will our particular mansions hold together, keep out the rain, shelter us from the winds, and allow us to make a home for our families?

OK, what in the world does this have to do with OPML?

OPML is a language for assembling sticks and timber of knowledge into useful systems that can be shared with others, that can be improved upon by a community, and that can be published to the entire online world at almost no cost.

OPML is the architecture of a system of knowledge. An architect of a stick-built home expresses a vision in plans. She draws floor plans (top view), elevations (side view), and specialized drawings for electrical, heating and ventilation, and so on. Similarly, OPML structures, as we will see, are used to assemble valuable pieces of knowledge that are wildly scattered across the web. Just as architects of buildings constantly consult books of materials and catalogues of lighting fixtures, heating systems, roofing materials and windows, architects of the web are masters of locating rich sources of potentially relevant knowledge. The artistry of OPML is to combine sources in an appropriate and elegant manner.

The biggest institutional supporter of OPML over the past several years has been Harvard University.

It is perhaps not surprising that a university would see the potential of an outline-oriented approach to online information. Outlines are a fundamental way to express knowledge.

The generality and power of outlining is most obvious in annotated bibliographies, which are probably the most comprehensive units of knowledge created and shared in an academic community. The libraries, in turn, are organized to help students, faculty and researchers generate such bibliographies and access the referenced materials. Research surveys, summaries of current issues, scholarly papers and books all have a fundamentally outline structure. Indeed, the university itself is structured as an outline: Teaching and research in a university revolves around outlines: the university and its departments, research centers and their groups, majors and concentrations and their required course.

Outlines in turn are portals to other forms of expression. One can have an outline of mathematical equations, an outline of field research, an outline of experimental data, an outline of audio and video recordings, or an outline of Second Life communities.

Indeed, outlines are a central structural metaphor for the university. And outlines are almost always how collaboration is begun (”hey, take a look at this outline”) and how the creative results of collaboration are summarized and shared (”here is an ouline of what we came up with”).

Harvard has long wrestled with how to use information technology to make its knowledge more accessible, both within its own community and, increasingly, with the rest of the world. For the “rest of the world” aspect, see for example today’s post below. There will be a lunch tomorrow at the Berkman Center, with Charlie Nesson hosting a discussion a conference scheduled for next spring entitled “How Open Will Harvard Be to Internet and Society.”

As part of this overall theme, Dave Winer was invited to be a Fellow at Harvard several years ago, and was asked to use blogging, RSS and OPML to help transform the University. If you want to get a rich sense of the times and the task Dave faced back then, read his wonderful piece for the Harvard Crimson, which is available online in blog form, with additons, on the web. This is a classic piece that deserves to be read widely. Here is my favorite excerpt. It is dated April 30, 2003:

I think weblogs are a very big idea. In fact I have a bet with Martin Nisenholtz of The New York Times, saying that by 2007 the top stories in world news will break on weblogs. This is not a merely a bet between gentlemen, there’s real money on the line. I’m sure I will win.

We’re returning to what I call amateur journalism, people writing for the public for the love of writing, without any expectation of financial compensation. This process is fed by the changing economics of the publishing industry which is employing fewer reporters, editors and writers. But the Web has taught us to expect more information, not less, and that’s the sea-change that the big publications face — how to remain relevant in the face of a population that can do for themselves what the BigPubs won’t.

Citizen bloggers in New Hampshire?

One of the best ideas I’ve heard so far came from Mike Clough, a foreign policy expert I met at Berkman. The idea is to somehow give a weblog to any New Hampshire voter who wants one, and then, much as I’m helping people at Harvard get started, we work together to help the citizens of New Hampshire get started.

Citizen bloggers covering the candidates for US president. Everyone who hears the concept goes Hmm, that might work. More than anything, I want the US presidential election of 2004 to be a real election, to mean something. I wonder if many other citizens feel the same way?

New Hampshire, so close to Cambridge, and with the technology so ripe, and the candidates so willing, it seems we may actually be able to route around the professional press and make something real happen this election cycle.

I’ll be visiting Dartmouth College (in New Hampshire) on May 9, and then will return during the summer, perhaps often, to interview candidates, and write about it on my weblog.

Our server is open

We’re just getting started with weblogs here at Berkman. We’ve opened a server, where anyone with a harvard.edu email address can create a free weblog. Our hope is that many people will take us up on this offer, and we can explore the potential of this new medium together.

Toward that end we have regular meetings every Thursday at Berkman, 7PM, see our weblog for details. Every meeting we spend about one hour reviewing the software, I answer questions, take requests, and drill the core stuff every week, so that newbies always learn something, and always feel welcome. Then we spend about a half-hour talking about what we’re learning and sharing ideas on how the technology might be better used.

Dave Winer

Thus began several years of experimenting which saw the first university-wide, open blogging service established at Harvard, and saw Dave convince the New York Times to become the first major newspaper to make available public RSS feeds. While both were Fellows at Harvard, Dave and Christopher Lydon did the first podcasts–which included realtime coverage of the US Presidential Primary in New Hampshire as well as upwards of a hundred interviews with experimenters with blogging around the world. Ultimately Chris took this around the world with web broadcasts from Africa, the Carribean, and South America.

Many of the leading applications of community information technology at Harvard are built on a foundation of OPML. Many Harvard bloggers have moved to outline-based blogging, using the OPML Editor or the OPML Workstation Writer and their associated free open hosting services. Complex, multiple-layered OPML outlines can be easily examined in the Grazr AJAX OPML viewer.

Professor John Palfrey at Harvard Law School oversees a comprehensive online resource on Internet Law in a public set of OPML files which combine course outline, syllabus, and RSS blog lists of experts writing in the field. The result is a stunning summary of the state of a fast-moving field, combining John’s overall perspective with daily, real-time updates from other experts in the field such as Larry Lessig.

John maintains the popular “Blogs for Teachers” public aggregator site, with OPML output, on TopTenSources.

The H20 online educational community, which is a multi-university collaboration to share course materials, uses OPML to make its library of course materials available in an open format. Users can make “Playlists” of educational materials and publish them in open OPML on the site. Anyone can visit the site and use the OPML materials.

Finally, OPML is being used by researchers at Harvard to conduct studies of communities of bloggers, literally around the world. Berkman Fellow Ethan Zuckerman of Global Voices recently used OPML outlines of Global Voices blogs to create a Google-based search service for the Global Voices community.

Indeed, the collected blog subscription information held within Bloglines forms an open dataset, in OPML, that can provide enormous insight into the reading lists of the hundreds of thousands of the web’s core users.

Companies that have been spun off into Cambridge’s OPML Alley include Mike Kowalchik’s and Adam Green’s Grazr, referenced above, as well as our own John Palfrey’s and Bela Labovitch’s TopTenSources, which in turn publishes human-and-machine-edited OPML source lists on many hundreds of topics. Pito Salas’ Blogbridge is the most powerful aggregator available, and allows remote access to OPML feeds–meaning that a complex network targeting shared, dynamic OPML feeds can be readily constructed and will operate automagically. In addition, Pito publishes OPML feeds for use by others, including his Top 100 blogs opml source. Last-but-not least citizen blogger and evangalist Lisa Williams’ (of H2Otown/ Watertown) soon-to-be launched Placeblogger site which uses OPML to provide central access to hundreds of citizen journalists in towns around the US.

———————————-

As I compose my book on OPML, I am of course influenced by what happens on the web as I write. I am going to include references where helpful to contemporary, topical material such as the announcement from the Berkman Center below. When I look back, these refernces will provide additional context for the book. In the present, these links and posts elaborate on the theme of the post.

Home - Berkman Center for Internet & Society
Tuesday Luncheon Series: Charles Nesson asks “How Open Will Harvard Be to Internet and Society?”

How Open Will Harvard Be to Internet & Society? That’s the big question for the Internet & Society 2K7 Conference, set for May 31 and June 1, 2007. What would a more open Harvard mean or look like for faculty, for staff, for students, for alumni? Is there an understanding among all as to what open access is? The process and the conversations in the lead up to the Spring conference are integral in the shaping of the conference; Professor Nesson will speak at this Tuesday’s luncheon at 12:30 p.m. about his hopes for Internet & Society at Harvard, but will as importantly be listening to the needs, desires and perspectives of those affected by the policies across Harvard and beyond.

Charles Nesson, with fellow Harvard Law Professor Charles Ogletree, has been chairing biannual Internet & Society conferences since 1996.

Charlie’s preliminary blogpost on this conference: http://cyber.law.harvard.edu/nesson/blog/?p=199
Bio: http://cyber.law.harvard.edu/home/charles_nesson

Join the conversation online:
Webcast: rtsp://harmony.law.harvard.edu/webcast.sdp.
IRC Chat: irc://irc.freenode.net/berkman.
Second Life: http://tinyurl.com/s6tv4.