Ross Mayfield's Weblog

Wiring the Wiki

The Wired Wiki project is really coming along.  The main article has over 100 edits from a diverse group, there are lots of headline suggestions and sanding of the deck.  So far, contributions have been constructive and civil.  Only exceptions so far have been from wiki vendors plugging or trolling (which is suprising given how they should, in theory, know the tool).

Today Ryan Singer started to share his Reporter's Notebook, so editors could source copy from his interview with Ward Cunningham.  We also split the Comments page with a Discussion page.

Edit this Wired News

Last time someone tried this it was a disaster, but Wired News has boldly put an article about wikis into a Socialtext wiki for anyone to be a Wired editor:

In an experiment in collaborative journalism, Wired News is putting reporter Ryan Singel at your service.

This wiki began as an unedited 1,059-word article on the wiki phenomenon, exactly as Ryan filed it. Your mission, should you choose to accept it, is to do the job of a Wired News editor and whip it into shape. Don't change the quotes, but feel free to reorganize it, make cuts, smooth the prose or add links -- whatever it takes to make it a lively, engaging news piece.

Ryan will answer questions from the comments page, and, when consensus calls for it, conduct additional reporting. If there's something he missed, let him know, and he'll get on the phone and investigate, then submit new text to the wiki for your review.

Readers can also submit headlines for the story, and write and edit the "deck" -- a blurb for our front page and RSS feed that promotes the article.

To make any changes, you'll first need to create a free account at Socialtext.

We'll release the results under a Creative Commons license, and, if the whole thing doesn't turn into a disaster, run the final story on Wired News on Sept. 7, 2006.

To clarify, all people have to do is register to edit, a spam countermeasure.  This is of course different from the LA Times experminent as there is a clearly stated goal.  It will be interesting to at least watch.

Socialtext Wins Bowling 2.0 League Championship

Who says social capital is dead in America, or people are bowling alone?

Tonight, in a come from behind victory, the dark horse candidate in sudden death, with the league watching every throw, we triumphed over Facebook. The turning point may have been when Led Zepplin's "Whole Lota Love" came on, and the kids couldn't relate. A noble victory led by team captin' and league leader Erik Langner.

In other news, Julia really won the Web 2.oohh contest, but like I said, bullshit.

Bad Joke Popularity Contests

The Web 2.Ohh contest for who was the hottie was of course a joke.  Looks like our very own Julia French has won, 8,647 to 814 at present count.  For the record, I communicated clearly on my internal blog not to script to cheat, and I'm not sure where such enthusiasm came from.  But this is bullshit on several levels.  For one, there is no prize.  For two, contestants are not public figures who should not be thrown into ring without their consent.  Three, cheap traffic trick.   Sure, it is all fun, but did I mention this is bullshit?

Hot or Not should be opt-in, especially when others profit from it.

10x 2.0

The traditional venture capital invests disruptive technologies that provide a 10 fold improvement and generate 10x return.  However, Web 2.0 consumer internet companies are not based upon disruptive technology, but potentially viral models of participation.  The consistent pattern is sharing control to create value.

Measuring the disruption in performance terms proves difficult and is an opportunity for research.  I have not found a way to measure the lower cost (zero marketing budget) of user acquisition and lower churn through virality and network effects.

But you can measure the impact of disruption.  Craigslist cannibalized the Bay Area classifieds market by $60 million by 2004, when they had $6 million in revenue.  This is asymmetric competition against models that provide less freedom to communities.  Consumer internet startups generally require 10x less capital to get off the ground (although not consumer, Socialtext was seeded with only $5k and six months of sweat, as an extreme example).  Realizing a 10x return with a developed community and functioning business model should be a layup.

But if this model requires 10x less investment, the lingering question is how $1B venture funds invest enough.  Or how much time even a small fund can spend with its portfolio.  The IPO window remains practically closed and burdened with $2M in new SarBox operating costs making it less attractive.  M&A is out of control and potentially furthered by the need for venture firms to exit earlier to make the model work.

We are in a bubble, that is representative of our times.  It began with Google's acquisition of Blogger, and it might have peaked with the rumored $2B valuation of Facebook (hinting at the cyclical tip from consumer to enterprise).  I'm not sure there is a lack of failure in this bubble, or if we just need to give it time.  There definitely are more plays in each category (e.g.  well over 200 in social networking or social bookmarking, possibly over 200 funded in video search).  I've said for a year that this bubble is build to flip.

The problem is there are too many companies being created that have no aspirations to be companies.  Most survey the feature portfolio of tier 1 and tier 2 acquirers and are precision guided towards flipping within 18 months.  Many make no attempt at generating revenue and most that do generate revenue from the advertising of other startups, let alone demonstrating a business model.

A dirty little secret for those actually building these things into businesses is that generating community value takes time.  Beyond the blog bump and 53,651, a community that stays with you and grows virally is tough to achieve.  Today's successful ventures are twice as old as the 18 month window.  They were born when it was a shitty time to create a company, were largely either hacks cast out openly to serve an immediate itch and community (e.g. Blogger, Six Apart, Technorati, Newsgator, del.icio.us) or almost accidental creations that caused a business model iteration (Flickr).

Today the problem is microventures optimized to flip are flopping, but they are too small to see.  It isn't just that VCs are taking the risk of the bubbles.  For every funded company, there are 10 with the sweat and tears of entrepreneurs that fade to black.

This is not all bad.  We are in the risk business.  There is creative destruction with tight iterations that provides a base of vibrant innovation for the economy.  But structurally, this bubble is attracting increasingly less experienced entrepreneurs (and investors) chasing the wrong incentives. 

Deletionists, Networkists and Blog Sourcing

The argument over whether to keep the Enterprise 2.0 article in Wikipedia continues, in this incredibly rich discussion page.  Bear with me here, as the innards of Wikipedia may not be clear cut enough for the blogosphere's preference for individual voices engaged in disappating debate. Samuel Klein, the Admin who undeleted the article, explains his decision and offers an opportunity for understanding:

I'm going to post a set of instructions for all of you bloggers, on How To Criticize Wikipedia -- so that you can do it productively if you want to.  Wikipedia is one of those rare communities where eloquence, discussion, and an idea about how things can be better can lead to an immediate improvement in process and content.

This experience has helped further my understanding of core rules (Neutral Point of View, Verifyability and Original Research).  The process of administrating them is decidedly non-bureaucratic -- this is not Kafka, this is made of people.  And to SJ's point, while process is an embedded reaction to prior stupidity, the openness of process execution and distributed decision making authority favor change.  Simple rules and open social interaction yield complex emergent behavior.

One man's complex behavior is another's philosophy.  Deletionism holds that the path to quality is through vigiorous defence of rules to keep articles out.  Inclusionism favors keeping and amending problematic articles.  User Pages tend to self-identify and associate with these philosophies.

In this debate, Deletionists have run up against what I'll call the Networkists -- a group of highly networked domain experts activated through blogs.  Most are only casual contributors, so far, and some have had negative experiences (and some for good reason) making reverted contributions.  If I had to offer a philosophy of Networkism, it is that the participation of domain experts in creating articles and debating their deletion increases the quality of the process, and most domain experts will not become core members of the community.  Now this is all fiction, but over time it will be easier for both Wikipedia to attract domain experts and for them to activate and self-organize, so there may be something to learn from here.

Many Articles intersect with networked groups, and in many cases those groups mean ill towards Wikipedia.  I see value in the role, not the philosophy, of a Deletionist, because practically they are the first and most persistent ones to debate against outside groups.  But unlike the Recent Changes patrol reverting vandalism, where domain expertise is not required, their activity directly excludes potentially valuable members of the community.

Blogs as Sources

A second issue which has arisen is the role of blogs as reliable sources, a guideline for verification and no original research rules.  While the guideline rightly suggests there is a difference between blogs (Reliability is a spectrum, and must be considered on a case by case basis...For example, the blog of an academic department is not merely a personal blog, but should be looked at in the totality of the source.) -- media literacy of the blogosphere is under-developed.  Having an institution behind a blog is a traditional fact checking measure.  Other sources are reliable because there is an editorial process prior to publication.  But with blogs, this process happens post-publication.  A blog or a post cannot be judged reliable in isolation.

With the Enterprise 2.0 Article, verification is provided by an article in the MIT-Sloan Management Review, an HBS case study, a BusinessWeek article and reference to a Don Tapscott book.  But for some, this is not enough.  So I leave it as an open question.  Should Dan Farber or Dion Hinchcliffe blogging under the ZDNet masthead qualify as a reliable source, or is what makes them reliable what links to them say?

Enterprise 2.0 Re-considered for Deletion in Wikipedia

The Wikipedia Article Enterprise 2.0 has been restored and is undergoing it's second vote for deletion (or keep).  You can share your thoughts on this matter on this page.

This is not really an opportunity for a bunch of bloggers to blather in support.  Given the attention the article will get, it is a great opportunity to create an even better article about an important trend.

For me, this has been a fascinating look at the part of the process that makes Wikipedia work.

Enterprise 2.0 Re-vote

After the Wikimania panel on Enterprise 2.0, I rolled up my sleeves and tried to improve the Enterprise 2.0 Wikipedia article.  All that activity caught the eye of an Admin doing his job, who excersized speedy deletion.  Little did I know the article had already been deleted, primarily because it failed to meet the neologism criteria. 

At first, I was a bit shocked, as my work disappeared.  It also took a bit of effort to understand the rules at play and try to picture the process over time and scale.  While away this weekend, a minor blogstorm happened around the deletion.  Dan Farber and Jeff Nolan have some great wrap-up posts and David Tebbutt captured some emails that flew around.  But as I like to say -- blogs are a sprint, wikis are a marathon.

Articles that already failed an Article for Deletion test is a candidate for speedy deletion which any Admin can do for good reason.  The appropriate response is to first for clarification by the Admin who did the speedy deletion.  You can also request a History Undelete, which does not require a vote, to build a case.  But ultimately, in order to have the article restored, it requires a re-vote.

I had the history undeleted for the Enterprise 2.0 Article here.  I invite contributions to help make it more robust, and draw in more reputable and diverse sources that may help it meet the neologism criteria.  I think it is important to give it time before following the process for a revote, but express your opinions on the Discusssion page.

At one point every term was a neologism, language evolves, faster for some people, and the new becomes mainstream.  What's interesting is that Andy McAfee's initial Enterprise 2.0 definition (freeform social software adapted for business) still stands, but is being extended in consistent ways.  Jeff Nolan's insight that enterprise mashups are of processes, while Web 2.0 mashups are more simply data, is significant.  Vinnie challenges us further.  Dion Hinchcliffe offers up a definition of Enterprise 2.0 (richer than what I quote) as liberation:

Enterprise 2.0 in general describes the liberation of often previously inaccessible corporate information to be opened up to general discoverability, consumption, and reuse using a Web-based model.

Remember e-Business?  When the web intersected with enterprise software, a new front-end was accessible across firewalls, consumers learned to interface directly (while being "managed") with transactions that went straight through the back office and across the value chain.  Now think about how the web, and its people, have evolved.  It makes sense that the web will reinvent enterprise software again.  Participatory models, mass collaboration and the economies of speed, scope and span mashups afford will make us re-think process itself.  Part of it is the liberation of information, but most of it is tools that don't get in people's way.  Which especially matters when they are trying to work together.

Plazes Goes Mobile

Took a break in Mexico this weekend, literally locking my cell phone in a safe and not bringing my laptop with me.  Still catching up, but I had lunch with Felix from Plazes today (disclosure: advisor) who told me I could have tagged the location while being safe with the new Mobile client.  The world is getting granular.

Also on the MoSoSo front, I added Twttr to the top right of my blog for those interested in the messaging of the mundane.

Common Attribution

One benefit of participating in LinuxWorld was spending time with peer companies.  A number of Commercial-Open Source companies have similar licenses to ours.  Because of the realities of networked use of applications, we share a very similar Attribution Clause to visibly give credit to the project if you use the code.  Attribution is a value consistent with the Open Source Defintion and as Creative Commons has demonstrated, attribution can foster innovation.  What I have to share is a couple of us believe we can look to harmonize our attribution clauses into what could become a standard.

No REST for the Wary

Over at the Socialtext blog, Chris Dent has a really geeky post for those with more APIs than business cards.  Our team has been playing with our new toys to create such things as a filesystem, a multi-wiki search interface or a Ruby on Rails client that was built in a couple of hours.

Office 2.0

Ismael Ghalimi is not only becoming the leading Office 2.0 blogger, covering small-scale productivity with zeal, he has turned it into a conference.  While I hope this doesn't mean that in the future everyone will be fameous for their own conference, the event has shaped up into something great.  I'm speaking alongside:

• Rod Boothby, Manager, Ernst & Young, [Blog]
• Jeff Clavier, Managing Partner, SoftTech VC, [Blog]
• Nik Cubrilovic, CEO, Omnidrive, [Blog]
• Scott Dietzen, President and CTO, Zimbra, [Blog]
• Konstantin Guericke, Vice President of Marketing, LinkedIn, [Blog]
• Kaliya Hamlin, Network Director, Planetwork, [Blog]
• Jason Lemkin, CEO, EchoSign, [Blog]
• Ivaylo Lenkov, CEO, SiteKreator, [Blog]
• Ross Mayfield, CEO, Socialtext, [Blog]
• Jeffrey McManus, CEO, Approver, [Blog]
• Oren Michels, CEO, Mashery, [Blog]
• Jeff Nolan, Evangelist, SAP, [Blog]
• Tom Snyder, President, iNetOffice, [Blog]
• Jason Roberts, CEO, Preezo, [Blog]
• Narendra Rocherolle, Principal, 83 Degrees, [Blog]
• Raju Vegesna, CEO, Zoho, [Blog]
• Charlie Wood, CEO, Spanning Partners, [Blog]
• Ray Wu, Director, Venture Management, HP, [Blog]
• David Young, CEO, Joyent, [Blog]

Not suprising these days that every speaker has a blog.

Hi, I'm the Prince of Azeroth. Let's Network

BusinessWeek has a funny ego piece about my playing World of Warcraft in their issue on competition.  Like Tom, I have issues with a game that provides negative incentives for real life and have paired back my playtime.  But the game has been rewarding beyond the game itself.

One small correction, Joi Ito was the one who brought together the We Know guild.  And the phrase Warcraft is the New Golf should be attributed.

Events this Week

We are playing host to Web Montag tonight at Socialtext, come by for demos.  Das ist gut.

Web Monday is an informal gathering (in and around Germany) aimed at bringing together developers, designers, entrepreneurs, venture capitalists, researchers, web pioneers, inventors, bloggers, podcasters, end users and other folks interested in web 2.0 (in the broadest sense). The goal is to better connect the German web 2.0 scene as well as improve transatlantic idea exchange between Germany and the US (and – hopefully – help bring some of those good Silicon Valley vibes to Germany).

There's that damn Bowling 2.0 thing going on tonight too.

If you are going to LinuxWorld this week, as I am, stop on by our booth (#215) for a treat.  If you can't make the megaconference, the LinuxWorld Wiki is here.

The Wiki is not a Dump Truck

It is a series of people.

Enterprise 2.0 Think Tank Session

Participating today in a think tank session on Enterprise 2.0 hosted by SAP.  When I asked if the event was [ x ] bloggable, they agreed, but also gave the ability for someone to say something is explicitly Not Bloggable.  Susan Duggan from SVWIC came up with a great way to express this, Nota Bene is the new Not Bloggable.

NB! blah, blah, redacted, etc.

The first session was a talk by Jeff Nolan, which is inherently bloggable. This will fundamentally change the way that we build and distribute software.  This is software designed for users.  To me, everything is networked so the idea of something being bloggable is fundamental.

If you were going to start a company today, to get VC funding it would be

• Web 2.0 in nature
• Built on the LAMP stack
• SaaS delivry
• Rely on "community" for development or adoption
• Have "r" in the name...Flickr, Zoomr, GTalkr, Taggr, Twttr, Wankr

Within SAP not much has been done officially, but there are lots of smart people looking at this.  No broad adoption of scripting languages, which is an essential component, but Craig Cmehil on SDN has some toolkits and there are great examples of these leveraging Netweaver, for actual company value.  We need to aggressively adopt scripting languages. 

SOA is dead, Web 2.0 is what matters -- SOA benefits are for people who are publishing software, Web 2.0 benefits are for people who actually use your software.

Mashups?  What's a mashup?  In the consumer world, you are really doing some data integration for a service.  But in our world, a mashup is doing process integration.

REST: Representational State Transfer, is HTML + XML without SOAP, with a stateless client/server protocol -- and essential building block for Web 2.0 when combined with scripting.  State doesn't scale to consumer level volumes.  Instead every message can be self contained.

SAP Challenges.  We really don't relying on other people's code.  We think software has to be complex -- if it is important, it has to be hard, and dammit people are going to have to go through training to use it.  Whole generations of users coming into our clients expecting something different. We don't harness network effects.  Turning your app into a platform is making it easy for people to build things you could never anticipate.  We will give you, the lead user, 80% of the solution and let you innovate on top of us.  Today, people need permission to do this with SAP.  Network enable your applications.  Enable data sharing and data defaults. Linkify everything.  Syndicate your content.  Turn your app into a platform.  Open up and build a viral social architecture. 

SAP has some amazing assets to leverage for our benefit, and we would be foolish not to do it.  Some really cool services apps have been built and prototypes (Duet, Muse, Torrent).  The volume of conversation about Web 2.0 is rising dramatically.  What if Zimbra wants to create a new experience on top of SAP.

There is a technical infrastructure, economic model, development ideology, licensing models and delivery model for the Web -- but little has to do with technology.

Ending open question: How does our business model get blown apart when we let people innovate and repackage our services?

Disclosure: SAP Ventures is an investor in my company, Socialtext.

Ethics Gone Wild

Paul Kedrosky takes a look at Girls Gone Wild.  Or rather, the actions of the entrepreneur behind it, and raises the question -- is there such a thing as entrepreneurship morality?

I tried to comment, but got an error, so posting it here (consider this a disclaimer for not writing more on what could be a very big subject).

Does business have a morality? No, but it has ethics.  Although you can get lost in definitions.

Social entrepreneurship is class of business with a fairly well defined set of ethics. It is also a investment thesis (pornotube.com is an ethical thesis to the contrary).

Enterpreneurship is like many professions without institutional barriers to entry (e.g. PR, and yes, Venture Capital, increasingly in the hedge fund age) -- ethics will be defined in industry groups, in some cases to gain advantage.  In most cases, explicit normative development only happens when egregious errors impact the community.  Then they act out of profit motive and perhaps before regulation is called for.

Enterprise Wikimania

Sunday morning I was on a great panel about enterprise wikis with Josh Bancroft, Ned Gulley, Michael Idinopulos, Karim Lakhani and Andrew McAfee. The panel was a true conversation, drawing in equal contributions from the audience. Andy's takeaway was that with a freeform enterprise wiki you can positively accept contributions because it is different behind the firewall:

Ross Mayfield said that in four years of building wikis for corporations Socialtext has seen precisely 0 trolls and 0 instances of vandalism.  I was astonished by this and polled the entire room.  No one reported even a single instance of counterproductive behavior on the wiki. 

As I've written before, one of the advantages the Intranet has over the Internet is that people within companies share a culture and norms, and are usually quite reluctant to overturn them.  In addition, vandals and trolls can usually be easily identified behind the firewall.  So perhaps I shouldn't have been so suprised that employees aren't using corporate wikis to act out.

The zero bad behavior (so far) can also be seen as an indicator that people are more often than not, good.

My takeaway was a general consensus that while the tool matters, practices make wikis work and we need to share more.  Socialtext shares best practices openly to advance broader adoption.  The audience asked for way for the conversation to continue, and for resources mentioned be available. 

Of course, we set up a wiki: https://www.socialtext.net/ewikimania

 

Web 1.0 Pitch

Victorious!
Originally uploaded by eekim.

Last night Eugene and I won the Web 1.0 VC pitch judged by Mitch Kapor, Brewster Kahle and Jack Herrick. Our pitch was a B2B exchange for trading bandwidth derivatives...but I think we won by speaking the loudest.

WYSIWYG and Wikis

Notes from a technical session at Wikimania on WYSIWYG.

Christoph Sauer questions if WYSIWYG is a good thing, based on his experience with a wiki within his technical university in Germany and creating WikiWizard.  Leslie Lamport (helped create LaTeX) in 1987 wrote a paper called Document Production: Visual or Logical (pdf).  Visual is you see is what you get, Logical is what you see is what you mean.  A Canadian study, "Are Wikis Usable?" (pdf) with 4th grade students showed the biggest issue (49%) was link creation and management (understanding hypertext). 

Wiki Markup cons:

• "This is not for me," or people have to learn how to use it.  They just know Word.  There is no explorative usage.
• "Text processing of the old days."  A "sea of monospaced letters," losing overview in simple text editors
• "Wiki markup mess" or no standard for markup

WYSIWYG was developed for printing a document or using the same word processor.  But what about sending an email?  Will people see it the same way you sent it?  What if you copy and paste content between wikis, CMS, documents...what if the CSS layout changes over time and what if your article is published and is effected by a medium changes, or your audience changes (disabilities). 

Frederik Brooks Prediction in 1986: We will not see advances of scale in programming until we seperate accidental tasks from essential tasks.  Transferring this to web authoring: the accidental task is formatting layout, the essential task is information + basic structure (emphasis and linking).  Wiki markup is part of the essence of a document -- teach the public!

Wiki Markup Pros:

• Concentrating onthe content.  For the author a clean seperaation of layout and content, no typographical errors. 
• Context sensitive display: reader and designer have homogeneous design.
• Speed.  The tool for knowledge workers, e.g. bloggers don't lose the "flow"  No time for formatting.
• Simplicity.  Allows easy evolving of wikis, e.g. scripting.

He gives a demo of an editor that is like the Wikitext mode of Wikiwyg.  You can use familiar tool bars to add formatting, and you see the commands to do so, so people learn Wikitext as they use.  But it also formats on the fly like WYSIWYG mode.  It also has a table editor and a outline view.  Suggests standardizing on three image sizes for image uploads. (Someone asked about adding license info for images, I suggested looking at Greg Elin's Fotonotes) It does a basic paste from Word function.

His core argument: we should have an alternative to both extremes (I agree).  A forward looking vision would be computer literacy that involves undersanding the difference between the model and the view, and keep extending punctuation.  He also wants people to work towards a WikiMarkupStandard.

I can't really disagree with Chris' findings.  Ingy once called WYSIWYG within Wikiwyg "training wheels."  I like WikiWizard's dynamic display, but wonder how big of a payload it is and think it needs more editing modes to be usable.  Our usability interviews showed a tremendous uptake in adoption from introducing Wikiwyg and believe multiple edit modes and simplicity of the editor make a big difference.

Heavy Lifting

One interesting Wikimania session by Seth Anthony presented some research on contribution patterns. My notes:

Only 10% of edits are high content edits. 30% of those are anonymous, none are by admins, 52% are by someone with a userpage, none have a barnstar.  The people who are creating content are relatively new, not versed in style guides and bureaucracy.  Their use of Wikipedia speeds up a little bit through use, but not much.

Admins, on the other hand, are relatively efficient in their edits and have a consistent pace.  They only edit within the article namespace 60% of the time.  In other words, most edits are revisions of vandalism.  But in the early days they edited less frequently, created more content (76% of the time) and edited less consistently.

Content creators seem to be occasional and less frequent editors that may be specialists (subject matter knowledge).  Admins are former content creators, now janitors.  One person in the audience volunteered that he just crossed over from subject expert to janitor to emphasize this point.

In other words, the core community within the Power Law of Participation, the 500 people that do 50% of the edits, or 0.5% of the registered population -- does the heavy lifting for subject experts. Advances like Wikiwyg may increase the number and diversity of subject expert contributors.  With 2 million edits per month and vandalism anecdotally scaling linearly with edits, janitors matter.  Recruiting core community members (not sure if his research showed the rate of conversion) and scaling the core in a non-bureaucratic way will be necessary for fulfilling goals for enhancing quality as well.

I should highlight, as with most things, Wikipedia is an exception for wiki communities.  Their committment to giving Permission to Participate results in a higher burden than most communities carry.  But even in an enterprise community, a core needs to arise to carry what some people would consider a burden.  Gardeners don't spend time killing varmints, but instead prune the information architecture and watering areas for growth. As a result, I'd wager that the percentage of high content edits is significantly greater.

Brewster Kahle on Universal Access to All Knowledge

An Impressionistic Transcript of Brewster Khale's talk at Wikimania...

You are really on to something. There is something big going on here and a lot of these talks are about trying to figure it out. At the beginning of the talk I'll talk about things I know about and at the end talk about things I don't know -- how is that for a VC pitch?

I'll talk about open source, open content and the rise of the technical non-profits. Universal Access to All Knowledge is a big goal, and if you accomplish it, what are you going to do? Move to Florida?

Wikipedia is the 15th most popular site on the web. This is because of the enlightenment goal. The goal isn't a technical one, but a structural one. Despite centuries old balance...1976 radical expansion of US Copyright Regulation. Property of IP is perhaps the worst idea since the Domino Theory. Information is knowledge, not property. Valenti's crowning achievement radicalized copyright regulation. Most people talk about 130 year protection, but it is the vast scope and repercussions.

First casualty was software. The response was Open Source licenses. MIT's sale to Symbolics, which forked development and RMS' experience lead to open source. This is Brewster's revisionist history, but it may be where it came from.

The second casualty was Music and Video. The response was Creative Commons licenses. Another response was organizations to facilitate community effort. We lost the help of institutions like MIT so we built new ones. The Free Software Foundation. DejaNews was a for profit, sold to Google, dissapated. IMDB, 6 guy community project was bought by Amazon. CDDB became Gracenote, Inc. WAIS Inc was sold to AOL. FTP Software sold to NetManage. Cygnus sold to Redhat. All commercial companies built upon community effort that don't last long. FSF is still around.

The response is the rise of the technical non-profit. Apache software foundation has no full-time employees, but is incorporated to last a while. OASF has gotten money not only from Mitch but from Foundations. Mozilla Foundation is a great success with Firefox and the Google toolbar (money) they spun off a for-profit company. Interesting ecology to watch and try and understand what it means. Linux. Internet Archive is based on the open access model -- can we get paid for the administration we do so everything we do can be openly accessible. Wikimedia foundation you know about. The rise of the technical non-profit is an interesting addition to the ecology, we went wrong with the over-corporatization post WWII. EFF., Public Knowledge and Open Content Alliance exist to enforce rights and serve us. We massively screwed up our law structure and the general approach of knowledge of property.

Open Hardware. Petabox, a cheap machine that is open sourced. The $100 Laptop Program has interest in the order of 5-10 million. What would happen if the next major laptop company is a non-profit? It is because they are non-profit that they are trusted and base on open work.

The structure is now in place to proceed towards Universal Access to All Knowledge. We have institutions dedicated towards these goals, but how are we doing towards it?

In Text, getting the 26-28 million books in the Library of Congress. 1 megabyte for a book, 26 terabytes, $60k cost for the entire library on a Linux machine. But I actually like books, the printed page. Created the mobile bookmobile, which has printed a million books. The cost is a penny a page, a buck a book means you can give books away. In our first debut of this was the supreme court when they were arguing to extend copyright another 20 years, but we lost that one. Erik Eldridge has one, two in India, one in Egypt, one in Uganda... this gets closer to universal access, but what we realized is we need to scan more books. One way to do this is send them somewhere else. The Million Books project sends them to India, but we had to buy 100k books to send to them, but not many others wanted to send books to them. So the Indians were scanning their own books, which may be the right thing for them to do. Put the scanners next to the books. Sending to India to scan is $10 per book, in the US it is $30. The automatic scanners are not effective, so we made our own scanner and can do it at 10 cents a page. Scanning 400 books a day. $750 million dollars to digitize the Library of Congress. About a year an a half of the LoB budget.

Books are within our grasp technologically. There are issues about if it will be done by non-profits or projects like Google Books. We have an orphaned works problem. The way you ask a question in the US is through a lawsuit, Khale and Eldgrige. But if you get to frame a problem (orphaned works) you have already won. Who would forget the orphans? Give the orphans a home!

Next is in-print works. Amazon is working the other way, from print to out of print. We have found with the Open Content Alliance something that works. Even Microsoft is giving us money.

In Audio, if you take all the published works, there are 2-3 million musical works. A fairly litigated area. Some precedent that ripping them and putting them online might not be okay. A lot of musicians just looking to be put on the internet. The Grateful Dead allowed people to trade music. The key was, as long as no one was making any money. This allowed people to feel good about it. Legitimate bootlegging copies by other bands. So we went to this community and said: "would you like unlimited storage and bandwidth for free." They said, "we don't believe you." And they didn't like lossy compression. We said try us. Got lots of Okays. 2k bands, 30k recordings, everything the Grateful Dead played. Many versions of each concert, as there are debates over microphone types. If you give something for free, not only is it not taxed, but you get a tax rebate. Getting Slashdotted is a nightmare, your ISP bills could make you sell your guitar or house. Europe has a different copy-write scheme for performances (50 years), so we are working with the Dutch government to make old stuff free.

In Moving Images, 100-200,000 films. Not much, makes putting them online conceivable. We want to do this with DVD quality, but we are finding lots of archival films that never had distribution. Have 30k films on the Archive, dwarfed by YouTube, which is cool. Discovering genres like Lego Movies. Lots of these things end up in closets. Putting them online is $15 per video hour. We will host it, if it generally belongs in a library and it is okay to share it.

Television, we have a big Tivo, captured a Petabyte so far of 20 channels over a couple of years. We made one week available, the week of 911, we put online a month after. We are now understanding in the US that the news comes with a point of view. Chomsky used to say you should read 7 newspapers a day, recently this might make sense. Getting multiple points of view.

Television is technologically possible, there are some rights issues, but we could do it all -- all text, music, movies and TV is within our grasp. We got a change in the DMCA, yea! But we need a lot of help.

Web. We are best known from our web collection, about a Petabyte in size. In the history of libraries, they tend to get burned, usually by governments, and then they are sorry for 100 years, but it is too late. The lesson from the Library of Alexandria is don't just have one copy. Give copies away. Our first shot at this was with the Library of Alexandria version 2. If we had six or seven of these around the world I could sleep at night. We are trying to do this through large scale swap agreements.

Here is Wikipedia in the Archive. But most people are using it to look at their own stuff, their old websites. One of the reasons this is working is because we are non-profits.

Books, Music, Video, Software and Web -- it is all possible. Some open questions if it is public or private, for-profit or non-profit. Is Google the only shot we are going to have at scanning Harvard's library? Looks like it.

I'm going to use this opportunity to advertise some projects we need help on.

Non-profit Open Networks like SeattleWireless.net or MIT roofnet. Telecom company interests are not aligned with an open internet.

Distributed ownership network -- SFlan mesh network.

Open and transparent Web Search System -- Nutch. Let's build some alternatives and be more creative. Recall which does time-based search on the whole Archive, a project done by one woman that indexed more pages than Google, then she went to work for Google and hopefully she will come back.

Privacy and Anonymity. It is now known that the US Government is monitoring us. Tor.

Defensive Patent License. What if you did a GPL for Patents? The DPL is a license that reflects a public commitment to defense, so our patents are forever defensive. Any organization may freely use these licensed patents while so publicly committed to defense.

An Open Textbook system, started by Wikipedia. The number one request we get for books is textbooks.

Add Attribution to Wikipedia. Gutenberg guys didn't were nervous about the copyright thing. We should know where the facts from Wikipedia came from. Go read about Transclusion with Ted Nelson, backpointers. Richard Feignman, a physicist in 1982, was talking about how many layers it would take from Propedia to Micropedia to books as sources.

Open Library: annotate the book collection. Why is this book interesting to someone in the modern world. What can we do to re-inject old books into today?

We can pull off Universal Access to All Knowledge. This is where Wiki is going towards, one of the great things that humanity will be remembered for, up there with a Man on the Moon in the mythology of humanity.

Podcast with Jon Udell

Earlier this week Jon Udell did a podcast with me.  Funny how just last night Ward Cunningham was saying the Jon is the tech journalist he most respects.

Wikiwyg

Given the mention of the Wikiwyg project by Jimmy in his talk, I posted my own mention at Socialtext.com.  My hope is that we help Wikipedia fulfill the goal of being an encyclopedia that anyone can edit.

Jimmy Wales Kicks off Wikimania

Jimmy started his opening keynote by showing the Wikiality video.  Today he is talking about Our Movement Past, Present and Future.  This is the most difficult talk he gives all year, because it is before people who know everything he knows plus more.

Wikipedia's mission: Imagine a world in which every single person is given free access to the sum of all human knowledge.

Lots of milestones over the last year for growth.  Seigenthaler Controversy: Apparently, there was an error in Wikipedia (laughter).  They dragged me on CNN to yell at me.  Just at the time when Wikipedia blew past CNN.com's traffic:

Nature article: for articles on selected scientific topics which were of the same length of Britannica.  Average of 4 errors per article in Wikipedia, 3 in Brittanica.  Wikipedia isn't rubbish, it is actually pretty good.  Our work is getting better and better, but we are also realizing some of the limitations of traditional works.  We got lucky.  We are stronger in science than other areas.  Reviewers wer told to focus strictly on errors, not style, etc.  Articles were of similar length, which meant it was some of our best work, rather than stubs.  We are not as good as Brittanica...yet.  One of the big themes of the next year is a turn towards quality.  We can be proud of our article numbers going up, but we need to focus on the quality of our central core topics.

The foundation is maturing.  It always has been true that the projects have been far ahead of the foundation's organizational capacity.  We are becoming a better run organization, with five employees.  At the point where we can finally start applying for grants.  Met Jimmy Carter, who explained that many countries in Africa only receive 20% of what the could but are too disorganized to do so, which reminds him of Wikimania. 

Introduced Brad Patrick, a new employee, who was a volunteer lawyer and is now general council and interim CEO.  Brad stands up and says he couldn't be happier about the trajectory they are on, but more than that, the great people in the community.

Wikia, founded by Angela and Jimbo, completed a round of venture capital financing after a long search for investors who were willing to meet my strong demands for support for the Wikimedia community.  A portion of the money they raised goes to the foundation.  Hired full time engineers to support the community, covering the boring bits.  MediaWiki is amazingly fantastic and driven by volunteers.  Will have a big impact on the software.  We have a total commitment to free knowledge and respect for communities.

Campaigns Wikia is a project to try to improve the political discourse.  We know, as Wikipedians, that wikis can generate a healthy dialogue and mutual understanding (when things go right!).  Hosting local meetups all over the world, with his travel schedule on http://world.wikia.com.

Announcements:
* Today announcing that the One Laptop Per Child project is including Wikipedia as the first element in the content repository.  We may see whole new languages in Wikipedia arise as a result.
* Wikiversity: a center for the creation and use of free learning materials and activities.  Create and host a range of free-content, multilingual learning materials, for all age groups in all languages.  Host scholarly/learning projects and communities that support these materials.  Host and foster research based in part on existing resources in Wikiversity and other Wikimedia projects. 6 month trial period and initially in 3 languages.

In the coming year, creating an advisory board.  Help with partnerships, public relations, financing, technology, administration

Wikiwyg is a joint venture between Socialtext and Wikia to bring WYSIWYG to MediaWiki..  Release date uncertain, but both wikia and Socialtext are devoting full time development resources towards it.  My anticipation is that this is a big part of wiki editing and free culture.  My friend who when to Harvard to study Chinese history spent the last 16 years looking at the 8th century.  Read about Wikipedia and was excited to contribute.  Looked up articles on what she knew (read like an elderly chinese man in san francisco who loved the old poets).  She never edited.  She clicked on edit and saw some scary markup.  Some think this was a good thing, a barrier to entry, but it keeps out some smart people, and it doesn't keep out idiots (they are highly motivated).  Wikiwyg, in some shape or form is the future of the internet.

Quality initatives: with more than one million articles in English.  We should continue to turn our attention away from growth.  WP:BIO  Our policies on biographies of living persons have become more refined and focus on higher quality.  Admins and experienced editors are taking a strong stand against unsourced claims, especially negative claims.  Typical pattern: minor celebrity sees something they don't like it, so the blank the article, and they may have been right but we don't handle it well and they have a bad experience.

Image tagging: huge progress last year on image tagging, with virtually all the images in some problematic categories eliminated.  Still a lot of work to do in refining and reforming "fair use."  We should limit fair use to some specific categories.  It is a good doctrine, but we will be using it more narrowly in the future as we push to get more freely licensed images as replacements.  I'd rather have an image from a Wikipedian than one with the limitations of fair use.

There is a commitment to roll out as soon as possible a "stable versions" experiment in German Wikipedia.  This has been under development, allows us to move further in the direction of openness.  Protecting articles during a rush of vandalism was always excessive.  Semi-protection is better.  Allows two goals: allowing someone to edit at any time, but letting the general public have a better view.  If we don't get this done and rolled out by next year, we are making a big mistake. 

Last year he talked about 10 things that should be free.

• Free the Encyclopedia: Wikipedia, mission accomplished (that phrase has been ruined, more like mission accomplished but we have skirmishes every day).  I propose that the foundation seek funding to hire community co-ordinators and recruiters for important languages where we currently do poorly.
• Free the Dictionary: Amazing work going on with WiktionaryZ.  Will probably start to be functional later this year
• Free the Curriculum: Wikiversity.  Proposes that we work with people like Taddy Blecher in South Africa to provide him needed teaching materials.

Others who covered this talk: Ethan Zuckerman, Drew,  Pete Kaminski, David Weinberger, Meredith, BoingBoing,  Cathy, Andy Carvin, Filipino Librarian, Jason Calacanis, Doc (can you tell the usual suspects showed up this year?)

Wikimania II

Great to be in Boston at the second Wikimania, the Wikipedia community conference.  Keynoting at the first one in Frankfurt was an honor and joy.  This time, I get to sit back and enjoy one of the most diverse hallways you will find. 

Wiki Wednesday in Palo Alto was a real hit.  Jeremy Ruston provided a demo of TiddlyWiki and talked through the unexpected use cases, from GTD and beyond.  Jeremy has something appealingly useful that is being extended through plugins by a burgeoning community.  Some of his core insights were about how real people don't use the web.  The AJAX-enabled accordion-style presentation-in-page UI is akin to tabbed browsing (which regular people don't do).  TiddyWiki is an app as an HTML file, which means he gets to joke with people that by visiting the site they have already downloaded it.  But by containing the app as just a file in a folder, installation is accessible. 

Pete, Adina, Jonas and I also talked about Socialtext Open and how improving the installer is our top priority.  A funny conversation ensued about if making install too easy meant you wouldn't value the outcome as much.  But it was great to meet some folks who did install and get their feedback.

Wikiality

I'm just getting off of vacation, but had to share this:

Apparently it took down Wikipedia, or at least it did if enough of us believe it.