How a takeover of Delicious by Twitter could prepare the web for the 21st century.

What can I say, I’m tired of big-ass internet corporations taking over my favorite internet platforms and hanging them on the clothesline only to let them die. I am equally frustrated by smart internet entrepreneurs who so eagerly sell off their creations rather than maintaining the faith and commitment to carry out their original vision.

I have been using Delicious since September 2004. My first bookmark was a post by Anil Dash announcing that the two lead singles from the new Postal Service album were available for free download. For the next two years “Such Great Heights” would consistently rank in the top five of all tracks on Last.fm. I was one of four people to bookmark Anil’s post. My next bookmark linked to a WordPress plugin written by Tom Gilbert that shows Delicious bookmarks on the sidebar of a WordPress blog, exactly how they appear today at the bottom right of this page. 163 others also linked to the page of the plugin.

But Delicious was much more than just a way for me to record and share interestingness from the internet. From 2004 – 2006 “Popular Bookmarks on Delicious” was a daily routine. The most popular bookmarks on the website have always been heavy on web design and technology, but they also tend toward skills building and making it as a freelancer. I can confidently say that I learned more on Delicious Popular than in my four years of university. Many of my fundamentals about digital photography, web design, computer programming, video editing and even writing come from the bookmarks that I found on Delicious Popular.

More importantly, Delicious was the first website to allow users to describe and organize their content using non-hierarchical tags that group content for individuals, but also for larger groups. In this way communities of strangers began to form around particular topics, such as ICT4D (Information and Communication Technologies for Development). Flickr and Technorati soon followed suit, and years later Chris Messina adopted the convention on Twitter to create what today we call the “hashtag” on Twitter. All of this, apparently, led to a profound intellectual crisis in epistemology, our understanding of knowledge. It also lead to the coinage of a new term that is hardly muttered these days, folksonomy, coined by Thomas Vander Wal in 2004 to describe the new methods of categorization appearing on sites like Delicious, Flickr, and Furl (now Diigo). The value and future prospects of folksonomies were debated by information architect Louis Rosenfeld and social media enthusiast Clay Shirky. By the end of 2005 the term ‘Folksonomy’ even made it into the New York Times Magazine, with a note that officials from the Guggenheim and San Francisco Museum of Modern Art were introducing user generated tags into their online collections. No one thought more deeply about the role of tags in describing and organizing online information than philosopher and author David Weinberger. Back in 2004 Weinberger was already thinking about Aristotle’s knowledge tree and the “three orders of organization.” Three years later and he published his elaborated reflections in the book Everything is Miscellaneous. Today Weinberger is co-directing the Harvard Library Innovation Lab and can still claim one of the ugliest websites on the internet.


In fact, it was a presentation by Weinberger and his colleague Kim Dulin about their Harvard Library Innovation Lab that convinced me to finally publish these thoughts that have been in my head since it leaked that Yahoo! was looking to hand off Delicious. As Ethan Zuckerman explains, the Harvard Library Innovation Lab broadly seeks to explore the future of libraries. Specifically they are looking at ways to visualize information about books and how they are used. Compared to websites like Google Books, Amazon, LibraryThing and Goodreads, libraries have generally been god-awful at giving their users any information at all about the massive collection of information and knowledge that they house.

Why doesn’t my library offer me recommendations based on what I’ve checked out? Why can’t I add my own tags to describe the books I read? Why don’t I know how other library users in my community rank the books they’ve read? Goodreads shows me the most popular quotes from the books I read and my Kindle shows me popular passages of books that have been highlighted by others. My library gives me none of this – it’s just a big building filled with a bunch books.

The three main projects of the Harvard Library Innovation Laboratory – ShelfLife, StackView, and LibraryCloud – aim to provide solutions to these deficits. They aim to give library users more information about the books they seek … and those they’ve never heard of. What are the most recent books to be checked out? What are the most popular books within any field or tag? How do books relate to one another? When is a book used in a particular class syllabus? What are the most popular quotes, excerpts and comments? What other languages is the book available in? Where are the best reviews of the book?

My initial reaction after listening to the presentation by Weinberger, Dulin and lead developer Paul Deschner was: about time. Finally, libraries might enter the 21st century.


My second thought was: it’s great that we’re finally making books more useful, but how much longer will books exist as we presently know them? Today most books are either collections of short stories or essays that have already been published elsewhere (usually online). Some books, like Kevin Kelly’s What Technology Wants are quite literally a collection of copyedited blog posts. Unsurprisingly, Kelly – the author of nine published books – says that it will likely be his last:

I suspect this will be the last paper-native book that I do. The amount of work required to process atoms into a sheaf of fibers and ink and then ship it to your house or the local bookstore is more than most of us are willing to pay anymore. And of course the extra time needed upfront to print and transport it is shocking. This book was finished, designed, proofed, and ready to be read four months ago. But atoms take time, while bits are instant.

Despite the fact that many of my friends are publishing books this year1, I think that we’re going to see a decline in the number of printed books and an increase in the type of essay-length pieces which – as Anil Dash rightly points out – elicit massive commentary on Twitter, Facebook, Digg, Reddit, etc.


Which brings me to my last reaction to Weinberger and Dulin’s presentation, and to my proposal for Twitter and Delicious.

Immediately after I sighed with gratitude that libraries might finally enter the 21st century, I also realized with consternation that the same problems facing libraries are also facing the World Wide Web. Just substitute webpage for book and all the same problems emerge: 1) we don’t know when a web page is included in a class syllabus, 2) we typically don’t know if a web page has been translated into other languages, 3) we don’t know if there are reviews of the web page unless links are left in the form of trackbacks or comments, 4) we typically don’t know how our friends (or strangers) rank a piece of content, 5) we often don’t know how one web page relates to another, and 6) more often than not we no longer see the conversation that the content of a blog post inspires (on Twitter, Facebook, etc.).

Delicious could have easily become Twitter (not to mention Tumblr) if only Yahoo decided to actually do something after acquiring it. It had all the pieces in place – a large user base that quickly posted small pithy messages as they came across interesting content on the web. Only it wasn’t social enough and it didn’t go mobile as we did. But it still has a thoughtful community and a valuable database of how internet users have annotated the internet over the past six years.

Delicious Twitter
Tags Hashtags
Network Bundles Lists
Link Shortlink
Delicious Popular Tweetmeme
Title/Notes Text
Title/Notes Highlighting*

On Delicious we use tags to organize our content. On Twitter we use hashtags. On Delicious we follow groups of users through “network bundles” such as my network bundle for the G5.

On Twitter we usually add a one- or two-sentence description or observation about the content we are linking to. On Delicious I see most users (myself included) do one of two things: either 1) add a one- or two-sentence opinion or 2) include a relevant excerpt from the web page, the equivalent of highlighting on a Kindle.

My proposal, therefore, is to merge Delicious and Twitter, to maintain the 140 character limit for text, and to introduce a highlighting function that allows users to emphasize the most important/relevant/popular passages within any particular web page. Such a merger would resolve the greatest deficiencies of each platform; it would make Delicious more social and it would make Twitter king of the annotated web.


The Annotated Web

On any modern browser I visit a website and I can press a RSS button to subscribe to the most recent content from that website. But there should be another button on my browser to provide me with all sorts of information about the webpage I am looking at. Information such as:

  • Relevant Twitter messages ordered by number of incoming visits
  • Facebook messages where content is publicly available
  • Highest ranking comments from sites like Digg and Reddit
  • Citations from academic journals
  • Class syllabi that include the webpage
  • Human translations of the text in other languages
  • Statistics from Alexa including traffic stats, search analytics, and audience stats.

There are in fact a number of bookmarklets and browser plugins that do all of these things, but they all exist independently. I have yet to find a useful and comprehensive overview of annotations, commentary, and statistics about a single web page. Back in 2005 Pluck.com launched Shadows, an attempt to do exactly that – to create a “shadow” over every web page that provides more context and information related to what you are reading. A couple years later it quietly folded.

Now there are signs that Twitter is starting to pick up where Pluck left off. In April last year it bought out Bit.ly and added t.co as its own link-shortening service. Months later it took over the “tweet this” button from TweetMeme to create a stronger link between webpages and the conversations they generate on Twitter. Most major websites have adopted the button. Twitter’s official Safari Extension provides you with information about related Tweets and the mentioned Twitter users for any web page. Today over 25% of Twitter messages contain links. Adding Delicious’ archives and user base would provide Twitter with an impressive layer of commentary, context and opinion floating above the internet as we know it. It would also help fix Twitter’s frustratingly ephemeral memory. (It is nearly impossible to find a Tweet from over two weeks ago.)


There are a number of proposals for what to do with Delicious. One group has raised a grand total of $252 dollars to buy Delicious from Yahoo and create a “community-run” service. Others ask Yahoo to open source it, though there is no mention of whose server it will run on. There are a number of great discussions on Quora about Delicious. (I have Quora invites if anyone would like one.)

I believe that Anil Dash’s distinction between “short content versus long content” or “comments versus ideas” or “annotations versus content” are false distinctions, but that they are very useful false distinctions. Today my own usage of the web tends to bounce back and forth between each. I spend around two-thirds of my time reading long pieces that I have saved to Instapaper and one-third using social tools like Twitter, Delicious, and Facebook to share information with my own brief comments. The challenge for Twitter – of whatever other group takes on the task – is to bring these two facets of the world wide web together while providing greater value to each.

1) In terms of delivering ideas to readers it makes no sense to publish paper books these days, and yet a good number of my friends are doing just that. Earlier this month Evgeny published The Net Delusion, more or less an extended compilation of his online writings about slacktivism, the involvement of western corporations in overseas censorship, government online media manipulation, and the failures of the State Department’s “21st Century Statecraft” policy. (Don’t worry Evgeny, a review is coming later this week.) Next month Daniel Hernandez’s debut book Down and Delirious in Mexico City will be published. The majority of the book – focused on Mexico City’s subculture scenes – is based on writings from his excellent blog. In May Cyrus will finally publish his The Internet of Elsewhere, a book he has been working on for years now that compiles much of his writings about the role of the internet in Iran, Estonia, South Korea and Senegal for Slate, Wired, The Economist, Foreign Policy, and NPR. Later this year Rebecca MacKinnon will publish Consent of the Networked, which will thread together the major themes she has been dealing with on her blog and elsewhere over the past few years. Global Voices other co-founder, Ethan Zuckerman, will also publish a book sometime in 2012 based on issues he has been wrestling with for years on his blog: “The ideas will be familiar to many of the readers of this blog – the book is a chance to explore ideas like cultural bridging, pervasive translation, structured serendipity and xenophilia at length.”

So why are all these friends re-publishing their ideas in book form? I have two hypotheses: 1) it’s a great excuse to take a bit of an intellectual sabbatical to piece together all the ideas and fragments of ideas that we struggle with on a daily/weekly/monthly basis and 2) the title “published author” still opens helpful doors. Essays and blog posts can get you speaking gigs and fellowships, but if you want a tenured professorship or to direct a research center you probably need a book or two under your belt. I think this will change over the next couple decades.