Technology: March 2006 Archives
This is just too cool. Brian Eno and David Byrne are planning to release the raw tracks from their hugely influential 1981 album, My Life In The Bush Of Ghosts, with Creative Commons licenses, allowing for their use in remixing and mashups.
The tracks aren't on the site yet, but I'll be waiting!
Thanks, Xeni!
Technorati Tags: music
I just got a really nice phone call from John Kenagy, CIO at the Oregon Health Sciences University, letting me know that the UW's Catalyst WebQ online quiz and survey tool, has won the Joanne R. Hugi Excellence Award presented by the Northwest Academic Computing Consortium. The Hugi Award "is to recognize and share information about outstanding IT practices among the higher education institutions of the Pacific Northwest."
WebQ is one of the great group of Catalyst Web Tools that Tom Lewis and a small crew of intrepid perl programmers write and make available for folks here at the UW. It's a great survey and quiz tool and it gets used for all sorts of things - from research surveys (it's been approved as a tool for human subjects research at the University) to student government and faculty senate voting.
In our brief conversation, John mentioned that NWACC doesn't like to give the Hugi award to huge megaprojects, but to "the things that quietly propel us forward", which I thought was a lovely way of expressing the kind of value that we get from tools such as WebQ. It's another way of expressing the value that Dave Weinberger famously stated as "small pieces loosely joined."
As part of the award someone from the UW will be talking about WebQ and the Catalyst Tools at the upcoming NWACC annual conference in Portland in June.
Congratulations to the whole Catalyst crew for a great tool, and thanks to NWACC for the lovely bit of recognition!
A good quote from Terry in a meeting I'm in:
"Troubleshooting in the 21st century is the art of correlating seemingly unrelated events."
Technorati Tags: quotes, troubleshooting
This is really interesting - Amazon has released S3, which offers open web storage for $0.15 per gigabyte per month, accessible via web services (REST and SOAP interfaces provided) and BitTorrent. Authentication is via Amazon login, and you can use access control lists to control access to the files.
Interesting that there's no WebDAV interface to S3, at least from my brief look at the documentation.
They're guaranteeing 99.99% ("two nines") reliability.
This could point the way towards ubiquitous cheap online storage accessible via Web interfaces.
Technorati Tags: web-services, storage, web-services
One of the problems with the way my feed aggregator (Bloglines) lists my many subscriptions is that they get presented in alphabetical order within a folder. There's now 63 blogs in my "blogs" folder. When I start to read I usually start at the top of the list, but rarely get to the bottom. Which means that I almost always read the Apple Blog and BoingBoing, and fairly frequently make it as far a Jon's Radio and mamamusings, I don't get to Susan Crawford's blog as often as I should.
That's something I intend to correct, because Susan is writing some of the most intelligent commentary on the state of the discussion of Internet law and regulation. Last week she posted a great essay on network neutrality, or as she thinks we ought to term it, substrate neutrality.
You should stop reading my blog and go read Susan's now.
But when users say "internet" they mean relationships. We forget, because so many machines are involved, that the internet is a social world. Users don't think about transport -- they're indifferent to the substrate. They care about what they do there. And what they do is create a complex adaptive system unlike any other communications network we've ever had before. The unpredictable ecology of the internet could never have been generated by a broadcaster or a newspaper. It's constantly revising itself in response to the feedback it's getting from everyone. And its value is almost wholly unrelated to the work carried out by the access valves, the gatekeepers to internet access. As I've said before, the internet is like an ocean, but formed through attention rather than nature. (And, just as we're almost totally ignorant of the life-forms beneath the waves, we don't know all that much about what's going on on the internet.) The essence of our relationship to this ecology, this complex adaptive system, is one of explanation/comprehension -- at the most. We can't predict what it will do next.
The point about this ecology is that it is largely indifferent to the substrate it's carried on. The CD is not the song. The term "network neutrality" doesn't capture this -- we should consider using "substrate neutrality" instead. Otherwise the network providers' arguments are so easy: "But it's our network!" they can say. (AT&T's new slogan is "Delivering your world," as if the online experience was a visual pizza. We won't even need to rise from the couch.)
Technorati Tags: network-neutrality
At last week's calendaring BOF at Etech, folks asked how CalDAV treats recurring events. I had to embarrassingly admit that I didn't know the answer. Just because I'm evangelizing calendaring interoperability doesn't mean I've actually read the spec! (he says sheepishly).
So for those who want to know, from the latest draft of the CalDAV spec - the answer is that recurring events are treated as a single object in CalDAV:
Recurrence is an important part of the data model because it governs how many resources are expected to exist. This specification models a recurring calendar component and its recurrence exceptions as a single resource. In this model, recurrence rules, recurrence dates, exception rules, and exception dates are all part of the data in a single calendar object resource. This model avoids problems of limiting how many recurrence instances to store in the repository, how to keep recurrence instances in sync with the recurring calendar component, and how to link recurrence exceptions with the recurring calendar component. It also results in less data to synchronize between client and server, and makes it easier to make changes to all recurrence instances or to a recurrence rule. It makes it easier to create a recurring calendar component, and easier to delete all recurrence instances.Clients are not forced to retrieve information about all recurrence
instances of a recurring component. The CALDAV:calendar-query and
CALDAV:calendar-multiget REPORTs defined in this document allow
clients to retrieve only recurrence instances that overlap a given
time range.
Technorati Tags: calendaring, caldav
Interesting - Dave Winer has taken my notes from Tim Bray's Etech talk, where Tim listed some specific problems with RSS 2 that caused the creation of the newer Atom protocol, and turned that into a list of things to avoid in doing a successful RSS 2 feed source, which Dave calls A Busy Developer's Guide to RSS 2.0
Technorati Tags: atom, RSS, syndication
I cut out from Etech before it ended, to get back to Seattle in time to see my son in a skit at his school tonight (gotta set some priorities, after all).
Some thoughts on themes that permeated throughout the conference:
Open APIs - Online apps are opening up web service APIs to the world, allowing people to create new combinations of services in creative ways. These are almost all web service APIs, some easier to use than others. This is being done by both the big guys (Yahoo!, Google, Microsoft) and the startups (Technorati, Ning, EVDB). As the Yahoo t-shirt and sticker said - "Mashup or Shutup",
AJAX interfaces - Ajax has arrived big-time. Offering more immediate feedback by allowing changes in web apps without round-trips to servers is a very powerful technology that resonates with users. Client-side Javascript is the engine that powers this, and has gathered a large following in this community as a very powerful language in its own right, because of its dynamic (lisp-like) properties.
Syndication is happening - while the question "Is the content in your aggregator more important than the content in your email inbox?" wasn't answered affirmatively by the majority in the crowd, the mere fact that it was being asked is indicative of the amount of information exchange happening via RSS and Atom. Between this and the open APIs for mashups, we need to adjust to a reality where we're not in control of the context in which people see our content.
Learning from gaming - We can now think about taking what makes games compelling to so many millions of people to design applications that have that appeal with the same emotional resonance.
Coping with community - We need to evolve new and better ways of dealing with very large online communities, numbering in the millions.
So, all in all, how was Etech?
It was really good to get a chance to meet some folks I had only had online communication with (like Tim Bray) or hadn't met at all (like Robert Kay from MusicBrainz and Matt Pasiewicz from Educause). And it was great to get a chance to hang out some with Ted Leung, Tantek Çelik and Catherine Yang from Educause. I wish there had been some more small-scale unstructured, or perhaps more loosely structured, time - but that's hard to do in a thousand person conference.
There were some very good sessions - I tended to be drawn to the talks that had the highest conceptual content, as opposed to those that detailed working code, perhaps because I don't spend a lot of time working with code these days. Linda Stone, Clay Shirky, Amy Jo Kim, and Danah Boyd were real standouts for me.
I felt good about going down and flying the banner of calendaring interoperability work - it's important work, and people were interested. I look forward to next year - maybe then we'll get a calendaring presentation on the agenda!
Update - Julian Bleecker wrote to note that he enjoyed the notes but I got his last name wrong - sorry, Julian!!! That's the result of trying to take notes from the very back of the room while not having a conference agenda handy.
Looking at the way physical environments become computational grids - physical structures for play in the real world in which different kinds of social groups and formations come together and engage in playful activity.
The way the social network leaks into physical space - not just the Internet, but telephony an dother technologies.
Paths are the interface - the way people navigate the city - the city as a game board grid.
Teh things that we see around and turn away from can become pieces of the game expereience - Debris Become Legible. These can become instcription devices for game play. The way infrastructure becomes part of the game - goals or pieces or moments for power-up. different aspects of urban space become a way for the game to pervade physical space. Annotation rewrites the rules.
Movement = Power-Up. Traditional console gaming doesn't address physical mocement. Shows a room full o fguys staring at a screen - this iw weird - social play should have another register of interaction and engagement besides sitting still even when you're in a group of friends.
Pervasive games are ways of experimenting with social contexts and groups of people. Look at different ways in which social beings can be mustered to look at things in a different way.
So What?
The stakes are about a different way of seeing the world and how it works - hopefully in ways of making the world more inhabitable and sustainable. e.g. seeing debris as something that needs to be dealt with.
The "Big Urban Game" Large totems that were carted around the city - semicodes were the goals fo the game - got points for retrieving using mobile phones. Marketed for Qwest - ConQwest - getting people to navigate physical space and encountering poeple not involved in the game, and that interaction is interesting.
Superstar - at UbiComp in Tokyo last year - put stickers all over Tokyo, have other people find and take picture of them, and send them in to the "mothership" - the more pictures you have the more points you get. A cooperative game - you want to link to others and have them take your picture at the same time. One of the goals is to enlist other people and get them involved in the game play.
Pervasive Performance - games that are stage as a form of urban play - Blast Theory group in the UK. Create a mixed-reality experience - one level is online play And then there are real people in the world (the performers) A PDA equipped with a GPS and radio - the online players steer and guide the performers to a goal, through a 3d immersive interface. The physical players (who are actors) gather a group of people who follow them around.
Ludic component - Geocaching - a casual gaming experience. An interesting way of combining people who participate (the cachers) and the people who search. It's a global thing - use of the whole world as a kind of gameboard.
The Go Game - The object of the play is to do insane activities in physical space - turn the world upside down. Gets the players "out of themselves" in a whimsical experience. It also turns the world upside down, as you see activities in physical space that you woldn't normally see.
Mobile Phone Games -
Using mobile phone can become an interface not for games on the screen, but playhing games that pervade the physical world. clickr! co-located individuals interact with an experience. You don't need the latest, greatest, phone to do mobile phone games. Participating in largers social contexts.
Flirt Stampede - uses cell tower location to create a virtual stampede. IC you're closer you'd see the stampede on your phone screen.
Flirt Lost Cat - a lost cat wandering around the city - depending on where you were you might come across the cat. Using our movement through the city as a way of creating a casual, playful experience.
Twitcher - using your phone as a way of capturing birds that are flying around. Your phone buzzes - you need to capture a picture of the bird before it flies away. If you're standing next to someone else who has the game, you might both have the opportunity to capture the first picture, if you're within bluetooth range.
Viewmaster of the future - Using quicktime VR and a sensor for orientation sensing. You look into it and as you pivot around the view changes. You can think about ways that these kinds of experiences allow you to experience a cinematic or game moment as you move about the world.
Human PacMan
Catch Bob - Using gaming as a way of asking research questions - how does collocation inform social interactions? Identifying where other people were and trying to attain a shared goal on tablet PCs as you played the game.
Deeing Yoshi - depending on pervasive network doesn't work, so try to use the spottiness of the network get used as part of the game mechanics? Food that Yoshi can eat are identified by specific wifi nodes - as you pass them you get the food - so as people played the game over weeks they'd find where the good food was - sometimes the node was doen, etc.
piedimonsters - they have courseware called service design. Interested in nutritional fitness. Idea was to turn pedometers into physical game by combining it with tamegochi. Tying nutrition and physical activity by putting it into a game. What kid wouldn't want to power up their avatar by walking to school instead of getting a ride?
research.techkwondo.com
TrimPage Junction Framework
NumSum - a web-based spreadsheet. A single page app. Allows saving offline, through Save Page As in the browser.
Next Action - GTD to do list app. Nice code viewer in the app.
Persistence Technique 1
Modern browsers keep dhtml DOM tree intact during a File Save Page As
Keep you data in the DOM tree
myHiddenDataDiv.innerHTML=bigString
Whenever user saves the HTML page, you're ok
Might not work in Safari
Persistence Technique 2
Flash 8 Storage
Flash to JavaScript bridge
Seamless!Except, when you hit squantum level storage usage (Brad Neuberg)
Persistence Technique 3
IE'isms
IE persistence
IE offline data
(but nobody uses it)
Persistence Examples
DOM Page Saving Tehcnique
- IddlyWiki & Friends
Num Sum
Next Action
Flash storage technique
AMASS demos, Tiwywiki.
You need Synchronization in addition persistence
- can use data/record level semantics, track deltas, change requests, not changes; INSERTs only; unique ID gen. OR just punt
You also need a client side API - VB style? No - Rails Style? You can get tw write once run anywhere. Do do that you'll need SQL on both sides - we know it on the server, but what about the client? Will Firefox have something?
TrimPath Junction
A MVC Framework for JavaScript
raiels-like API with client-side SQL
Designed for write once run anywhere - server runs Rhino
Designed for pluggable client-side storage.
eval and with in JavaScript - changes dynamic scoping.
Why care about with?
Domain specific mini languages are easy - JSP, ASP, SQL are examples of mini-languages.
HST Templage engine
JST==JavaScript Templates 297 lines of code.
TrimQuery SQL Engine - RexExps to transform SQL to TQL
the problem -
Web 2.0 mantra:
- Apps run on server
- client is just a communication and display device
Ajax assumes constant connectivity
Connectivity may be flaky or slow or unavailable.
Leads to a style of work that can be termed Frequently Connected Computing
- a presumption of network availability
- bot both planned and unplanned outages occur
- you want to continue some work during disconnections
drive-through-internet.org
To do this, apps have to be disconenction tolerant and delay tolerant.
Hw to do this?
- put more of the interface on a single page
- do serious error checking
- don't lose user input
Reduce number of pages -
In disconnected mode you don't want to:
- change pages for routine operations
- rely on server data for routine operations
- fail mysteriously or verbosely on update errors
Leads to single-page apps, but with Ajax backend.
- Download "working set" with page
- this can be done with background ajax
- pace requests properly until page populated
- Structure UI, e.g. using tab-styled navigation
- don't forget bookmarkability (update location.hash)
In error checking there are bugs in browsers you have to work around - error-protecting the error detection. Keep in mind that responses may not come back in the order they were sent.
Don't lose user input - can be kept as page state (e.g. in text areas); JavaScript state (in variables) - both of those methods have problems if the browser or system closes. You can store it in cookies or in some storage made available by a plugin (like flash SharedObject).
PANIC - Persistency for Ajax in Networks with Intermittent Connectivity.
You need to plan for data changing on the server while you're offline. Use conflict resolution strategies, like server-based generation numbers an dlocal generation numbers. THe cookie size limitation is a problem.
Proof of concept implemented on Prototype library, server parts based on Rails before_filter.
Terry Gray sends along this tidbit (click on the quote from Marissa Mayer) from Fortune Magazine, by Marissa Mayer, Google's VP of Search Products and User Experience:
I don't feel overwhelmed with information. I really like it. I use Gmail for my personal e-mail -- 15 to 20 e-mails a day -- but on my work e-mail I get as many as 700 to 800 a day, so I need something really fast.
I use an e-mail application called Pine, a Linux-based utility I started using in college. It's a very simple text-based mailer in a crunchy little terminal window with Courier fonts. I do marathon e-mail catch-up sessions, sometimes on a Saturday or Sunday. I'll just sit down and do e-mail for ten to 14 hours straight. I almost always have the radio or my TV on. I guess I'm a typical 25- to 35-year-old who's now really embracing the two-screen experience.
ITags - something she and her colleagues are proposing to deal with some problems with tagging. She's started a company called Dabble, which is a video remix community.
There's a problem with technorati tags - people want to be able to publish their own stuff on their own sites and have people know that it came from their own sites. There are lots of different uses for tags and people want to participate in communities, but still publish on their own sites.
She did a usability study last spring, centered around text bloggers - what they asked for was the ability to create tags that didn't require a link. At same time they realized their categories were being treated as tags in aggregation systems. Tags needed to be trusted - wanted to know that tags weren't spam but didn't need to know the actualy maker. Visibility vs. invisibility. Want to create their own tag clouds, and to tag objects separately from blog posts.
35% of technorati tags are actually tags, but 65% are blog categories. In creating Dabble they're collecting media not just tagging it. They've collected 68,000 items so far - about half have tags. The richer the media, the more tagging you get. 10,000 of the videos come from independent sites with video on it.
What happens when people check out video is they look at the duration and the tags, much more than the titles.
What are itags - identity, creative commons license and an object. All of this can be bound to an object, which can move. The degradation of URLs is very fast, but the media lives on somewhere.
The identity piece - can be well-known, psuedonymous, or a blog url. You can protect the suer by obscuring who they are. When you trust tags (if you're an aggregator or host) you have to trust the maker of the material - so you need some sort of identity. This allows freeing of media from file location to be site-independent - could use XRI as a structured identifier.
http://itags.net/
A few years ago you could argue that RSS feeds and blogs were synonymous - that's not so much the case now, with podcasts, watchlists, commercial publishers, other peer-produced content and rich media. People don't realize it's rss feeds powering this dynamic information.
Full or Partial Feeds?
- Feed subscriptions to full content feeds outpace subscriptions to partial content feed for the same source by up to 10x.
BUT
- Feed subscriptions with only partial feeds grow just as quickly as subscriptions to sites with only full feeds
AND
- experiments in decreasing the amount of content in the feed to NOT statistically improve clicks back to teh site (title only)
SO
- Content providers can choose to distribute more or less content but the feed may be all that a consumer ever sees.
Complexity breeds consolidation, simplicity doesn't
- 2004 - Couple hundred clients pulling feeds
"My Yahoo will eliminate these other things"
"Desktop readers will go away - it will all be web based"
"soon there will only be two or three of these"
- 2005 - Thousand Clients
"Google will jum in here"
"soon there will only be two or three of these"
- 2006 - serveral thousand clients
- "when IE 7 comes out all these other things will go away"
- "It will all be personalized Ajax Home Pages with filters and aggregation"
"soon there will only be two or three of these"
Subscription without recognition
The Feed is Your face - more people will interact with you through your feed, not your web site.
Have content, will travel - you can't be control how your content is going to be redistributed.
Feedburner allows you to attach "beacon gifts" to your subscriptions so you can get an idea of how many people are viewing your feed.
Interactive language of attention.
Kathy Sierra talked about the power of the tribe - when you have a tribe you have attention. Putting the code libraries and APIs and patterns out is leveraging the power of the tribe, creating a tribal platform.
Opening the APIs internally has had a great impact in Yahoo - having hack days and mashup days internally to motivate and excite people. chaddickerson.com - mobile phone app to show location of your flickr group on the etech conference floor.
Two recent examples:
Yahoo! Design Pattern Library
Yahoo! User Interface Library
Design Tribe - pattern library. a new model - the page goes away and you have rich interaction without page refresh. But it leaves lots of things in a quandry. Engaging the user: Wow! gets their attention, but unless it's relevant (Delight!) it's not going to be used, much less inspire loyalty (Love!).
Vocabulary - interactive patterns of attention.
Immediacy - the Live Suggest pattern and the AUto Complete pattern.
Directness - Drag and Drop pattern. Inline Editing pattern.
Invitational - (being polite and inviting) hover invitation, tooltip invitation
Without boundaries - Endless scroliing, in context expand, hover details
Light Footpring - remembered collection, rating an object
Cinematic - fade transition, self-healing transition, sliding, fading
Rich Content - shareable object
They'd like to get a more common vocabulary around these design patterns - so they're starting to put them out to the public. Surfacing a vocabulary.
Exposing solutions - on the devloper side they've released the UI code library.
George Dyson gave a great exposure of the history of computing up through the last of the Von Neuman collaborators, drawing conclusions about how in the current age we are finally achieving the goals of Turing and von Neuman and the others in creating what is an intelligent set of machines - is Google alive?
The notes I took until I just couldn't keep up:
"Intellectual activity consists mainly of different kinds of search" - Turing
One of the arguments against artificial intelligence is that it can't be intelligent if it has ads - Butler thought in 1901 that intelligence would arrive from advertising.
The basic premise is that machines are both dumb and intelligent.
HG Wells thought of the World Brain, which is basically the Web, in 1938. He wanted it to be uncensored to prevent tyrrany.
Alfred Semee - Tried to figure out how the human mind worked electrically, in the mid 1800s. Wanted to build a search engine, but it would've been larger than the whole city of London.
Turing;s 1936 paper proposing a thinking machine was just as out there as Smee at the time - Godel argued that it couldn't doe things that the mind could, but Turing thought differently.
The real founder of the whole digital thing was Hobbes, who invented the idea of recursive functions. The bigger question is not whether the machine can think, but is it alive? Liebniz invented the binary calculus, and invented the idea of a shift register. Von Neumann took off from Liebniz and built the adders. Back then they let physicists build things - they didn't have Homeland Security and let Hungarians play with plutonium.
Exactly 50 years ago was the Von Neuman's first status report.
The machine was built at IAS at Princeton - On the first floor was Einstein, Veblen, and von Newuman. Godel was upstairs above von Neuman.
Up till then mathematics had been used to mean things, now numbers are used to do things.
Bringing engineers with soldering guns into the Institute was considered very odd. Von Neuman was interested in how to construct reliable machines from unreliable parts.
The real origin of the digital universe are tied up in the development of bombs, and much of the history is classified. Klara Von Neuman came over and learned to program the computer. Just a few weeks ago a filing cabinet turned up that has all of the letters between Klara and John.
Nils Baricelli created simulated organisms when the bomb guys went home at night
.
Ulam was working with Turing on multicellular processes - Von Neuman died in '57. His last book, the Computer and the Brain.
This conference has its feet in two paradigms - one is the attention economy, and the other is the old economy. Michael thinks we don't know which world we're in - butterflies (or moths) that still think we're caterpillars - but maybe we're ready to sip the nectar and live in a new way of being.
A new level in the game known as Western Culture
An economy most generally...
is a massively multiplayer
single level game
that involves some kind of passing
....?
Economic History is a multilevel game:
The first level: Fuedal (800-1200)
Knights and their vassals - winning through fighting, aligning with other lords, or by marriage.
Second level: Market-Money-Industrial - 1650-1980
Nothing changed externally (1200 - 1650 was a transition period)
Coming next and already very powerful - the Attention Economy - a new level, a new kind of system. Can't think of it in the old terms any more than you can think about stock markets or unemployment in a feudal system
When enoguh players reach their implicit goals, we reach a new level. Each level has new rules, goals, moves, values BUT - there is no game designer. The new level emerges from basic human proclivities and gaps in old level.
Feudal - provided security to Western Europe (from external raiders). Success in MMI is material abundance. The openings in feudalism ungoverned city spaces, safe travel, mostly no slavery. Our current openings - large high schools, broadcasting, publishing, internet.
Lack in feudal system was material goods. MMI lacks - chance to get attention and personal uniqueness and self expression.
Goals - fuedal: loyalty, fiefdomes, security
MMI - material goods, money, jobs
Attention - Attention from others
Roles: Feudal - Kinights, Serfs; MMI- owner, worker/consumer; Attention - star, fan
Different moves too.
Different cycles and structures.
So what is attention?
Attention - is scarce, very desirable, nothing limits the amount of attention you could absorb (if you could get it). There will be an intensifying competitiion for attention. Paying attention amounts to temporarily (and therefore permanently) allowing another to shape how your mind works.
It's not about time. Who owns your attention? You own part, but the person you're paying attention to also owns part. Attention is a kind of property located in the minds of those who have paid attention.
Finding meaning in life comes from sharing meanings with others - that can only happen if you get some of their attention.
They're announcing three things today:
Live.com new version
new version of Windows Live Search
Windows Live toolbar.
Live.com - launched in November. Wanted to give people rich control over information they view - what modules get shown on the page. Ajax based (of course).
The ability to add and name tabs (looks the new Catalyst web tools interface).
Does RSS aggregation - can hover over feed and see content from the feed.
Made performance improvements to live.com.
Search - can tab between vertical searches (images, news, etc) keeping the search term intact - like Google.
You can pop open a box to search within individual sites.
There's a smart scroll control that does dynamic fetching so you don't have to do individual page loads.
New Image search with some UI improvements.
Feed Search - can preview entire feed in search results before subscribing.
Search macros - ability to describe, save, and share a saved search with parameters. Can't yet create new macros, but will be able to in a few weeks.
Windows Live Toolbar - includes antiphishing filter, synced up favorites for IE, MS has acquired OnFolio - a browser add-in to help people organize info, including RSS aggregation and collections (gee, like Plum). Free with the toolbar.
Windows Live Local - being able to search for local info. Birdseye photo images - hired planes to take pictures. streetside - sent cars out with 10 cameras each to digitize images of downtown - Seattle and San Francisco so far.
The concept of Blue Chip product vs. the Off Bran product - Julia Roberts vs. Sandra Bullock, iPod vs. Creative Zen, etc.
How you could make blue chip products - the formula:
- Make people happy (by giving them control)
- Think about emotions
- Obsess over the aesthetics
Ajax's instant feedback makes people feel like they have control and makes them happy.
The iPod has obsession in making the entire experience seamless.
Some apps and how they live up to the formula or don't:
reddit.com - people share links and other people can vote on it. The reddit mascot alien is adorable - why is he cute? He has large eyes and he's bald, so he looks like a human baby, which we think is cute. The alien changes with little stories from day to day, which creates an emotional bond. Happy: A, Emotions: A, Aesthetics: B Final score: A
motorola Razr - best selling phone of all time. A lot of obsession over aesthetics. There's also the PEBL - people love to feel smooth ocean beach stones, so they made a phone that people like to fondle. Motorola gets an A for emotions and an A for aesthetics, but they get a D for happy for not giving user good feeling of control. Final grade: B
Web-based calendars - Basic requirement was being able to enter a flight which starts and ends in different time zones. Tried Airset - not beautiful, but very functional - even has a java application for cell phone which works well. The first actual usable java application for cell phones. A for control, B for emotions, C for aesthetics.
There are three web design trends that have become popular in the last year - ugly, ugly, and ugly. Enough with the pastel arial, already! This is copying from Google - which is cargo cultism. Google design only makes aesthetic sense compared to the old cramped up portals.
Look at architecture 9 the Seagram building or LEver house of minimalist design - revolts against Beaux Art wedding cake monstrosities.
When Google came out with the stripped down design, it was making an interesting statement. WHen you copy their fonts you're not making a statement.
The place I direct my attention when I have a question is to the search engine. There is intelligence on the web...and it's you. Great ideas, but can we rely on imperfect humans? What's the mechanism for doing that?
What if there was a way to collect almost anything from anywhere, and then share it in a way that matters to me, can I connect and discover things from other people?
so they created plum.
Lets you click on a page and plum collects it and caches it. You can collect anything - including people (within plum). You can collect live feeds. Other people can contribute to your collections.
You can collect email from your desktop or gmail or yahoo.
There's a little floating dock of icons called the plumber that has the controls for the desktop.
Today they're announcing (of course) an open source API. developer.plum.com
EVDB - Founded in 2004 to maximize event discovery.
Building a web services API platform and an event portal built on that platform.
The platform - REST-style API: api.evdb.com. Podbop is a mashup based on this.
The portal - eventful.com - a site to discover, post, and share events. Three dimensions to events -
Known events (search, calendars, tags, rss & ical feeds, groups, friends family contacts, auto-submit, iTunes import > calendar, AIM EventfulBot, Performers) everything tagged with hcal and hcard.
Expected Events - Prospective Search, notifications/alerts
Dream Events - events you wish would happen. tools for demand aggregation - not onlyh for fans, but for performers. Along with noticiations and alerts. ROlling out the first phase of this today - called eventful demand. Not just a "wish list" but the possibility of fulfillment. A marketplace for experiences.
They have some cool tools for creating and nurturing a demand for an event, including emailing people, and stickers you can put on your site.
Client technology for web applications.
We have Ajax, but there are some things we want to do - vector graphics, safe cross-domain scripting, storage better than cookies, etc.
This morning he's announcing the connection of flash player to ajax. Open sourced in AFLAX (the framework around ajax and flash) and dojo.storage.
These are all enabled by Flash Player. The flex framework is a way of typing in xml and script code to generate these apps. Flex/Ajax bridge. All these components are downloadable free.
He shows a demo of a flickr browser app that uses ajax and flash, with server logic in Rails.
He demos another application that doesn't use flash but has a data connecter component that updates pages and local desktop data simultaneously in more or less real time.
A pattern language is a tool adapted from architecture that's detailed enough to see how to build it, but not so detailed that it's tied to one domain.
Wants to propose a pattern language for moderation strategies.
Imagine a measurement of communal freedom - how much freedom does the software allow for users to communicate with each other as a group. Notepad is an example is one extreme, usenet is at the other.
Now imagine a y axis that is annyingness - flaming, trolling, unfunny cascades.
The problem - you'd like to launch apps that have a moderate amount of freedom with a moderate amount of problems - the reality is that once you cross the point of letting communication in you very quickly get into mitigating problems.
Slashdot illustration - over ten years has done a remarkably good job of not getting swamped by negative social effects - how do they do it? Members defend readers from writers. Users with high karma form a defensive membrane. But how do you design the system? Every comment is rated on a seven point scale. The members who moderate the system rank the posts. Most readers never see comments rated 1 or less - which is about 20% of the posts. Whether or not you can moderate takes four decisions. That's way complicated - there's very little use of the slashdot software by other sites. How can we get at that value derived from the knowledge of slashdot.
Slashdot faces the tragedy of the commons - each poster has an incentive to defect from the commons to get the most attention to their post. What slashdot does:
1 move comments to a separate page (reduces the size of the commons and lets people know that comments are somewhat ancillary)
2. Treat readers and writers differently - is able to defend readers from writers
3. Let users rate posts.
4. defensive defaults
You can imagine taking some of those systems and re-implementing them yourself.
Adding moderation system allows another problem - who will guard the guardians?
Another set of patterns:
1. Treat users and members differently
2. measure good behavior
3. enlist committed members
4. judges can't post
Clay is proposing building these kinds of pattern languages.
Bronze Beta (Buffy The Vampire Slayer fan site) made a set of different decisions, which show other patterns.
1. Don't Have Features (not an accident, but a strategy to reduce complexity.
2. Make Comments Central (unlike slashdot)
3. Make Login OPtional
Pattern Language Wiki and List - social.itp.nyu.edu/shirky/wiki
moderation_strategies@yahoogroups.com
Hobbes and Rousseau Argue about Dave Winer
in 2003 Dave Winer was running a mail list called blogrollers. One day he turned it into a moderated list, which was not well received by the members. The conversation around the change was about what should have happened - what duty did Dave have to his users? That's part of why we need a pattern for the long time.
Thomas Hobbes in Leviaathan - a monarch is required to keep society from descending into chaos.
Rosseau took issue with this - Force is no reason. People have a moral right to depose their leaders.
This is the context in which the conversation about social software is taking place. You can imagine that conversation going on, leading to the title of the talk.
Best line of the morning so far:
Social software is the experimental wing of political philosophy, which doesn't even know it has an experimental wing.
We are all seekers of attention - we all have ideas we want to promote. To that end we make claims on other people's attention - how can we reward that attention?
There are sets of patterns here - four patterns:
1 - heads, decks, and leads
The information architecture of newspapers - give people the ability to scan what you're presenting at multiple levels of detail and work their way into the material.
It's hard to write titles - or name anything. There's a cognitive dissonance between our objective sense of what we're about and what we project - we think people can read our minds, so why should we have to externalize ourselves?
What do you present the world in search results, for example? He shows an example of Google where the search results from a blog only shows the name of the blog instead of the title of the entry. Put richer metadata in titles.
The tragedy of discussion threads - a huge amount of collaboration goes on in mailing lists, and the title gets repeated on every message on a thread - what would it be like if discussion threads had titles that told a story? You could get a sense of the discussion from reading the titles. But it breaks the threading!
2 - active context
A new pattern: What do you know about a topic? The answer is most frequently a set of URLs. An active resource collection. A bunch of effects can happen - what tags are related, who are other people with this interest. Scoping is interesting - right now there's me and the world, but not yet let me see stuff by my trust circle.
Active collection is futureproofed in an interesting way. Used to put together a set of things in a list, but now send a url, which in some sense is a promise to keep updating. The token you hand someone is actually a query.
We can help people visualize what's changing in complex information environments - e.g. walking through the history of a wikipedia page.
3 - canonical names
OCLC ISBN relationship mapping as a mapping of relationships.
Soundbites in IT conversations - can create a url that is a specific time snippet from media resources.
4 - multimedia storytelling
Storytelling is what we do as a tribe. The ability to remember information we've gotten from listening or viewing by where it happened in time is profound.
Last night I moderated a BOF session on calendaring. We had about a dozen people show up, which isn't bad for an 8:30 pm session at the end of a very full day (and in competition with a bunch of receptions and parties).
Ted Leung attended from OSAF, which was good because there was lots of interest in Chandler and Cosmo and Scooby - folks seemed particularly interested in reusable AJAX-y sort of widgets to use in managing calendar data on the client side before the actual presentation is done. There was the predictable interest in how to work with Outlook/Exchange. There was also a lot of interest in the hcal microformat, which seems to be gathering momentum.
Lots of interest in caldav, of course. We talked up the standards efforts and the CalConnect calendaring consortium.
After the calendaring BOF I stuck around for the microformat BOF - by that point most everybody was getting punch-drunk, so there was lots of laughing and joking, but I got to chat with Tantek a bit about hcard, and he helped walk me through generating an hcard for myself and putting it up on the web - it's downloadable here.
Tim Bray - currently working at Sun, is talking on the Atom protocl as a case study. What lessons can we take away from Atom - was it necessary, was it done in the right way, and what is the right way to try and standardize Internet interop protocols.
slides at tbray.org/talks/etech2006
At the moment RSS is the most successful use of XML in the world - 27 million known feeds - 2 million feeds being updated every week. So why try to replace it with something new?
The problems:
Enclosures - the current RSS spec says you can only have one per feed, though many podcast feeds use multiple enclosure - clients vary unpredictably in how they support them.
Silent Data Loss - if you put in a string like "AT&T" that's not well-formed XML, the kind of markup you have to do to work in the different clients vary widely. What happens when you want to post something like "
Especially problematic if you need to put characters like angle brackets in feed titles.
Links - RSS feeds don't know about relative linking paths.
Problems - URLs that aren't all ASCII - like characters with umlauts. There's a standard called IRIs for this.
Problem with APIs - the MetaWeblog and Blogger APIs are under-specified, under-secured, poorly debugged, offer little interoperability, and omit many important authoring features.
So we have issues in RSS. We have a lot of experience with RSS now and where the problems are. THe normal course of affairs with the Internet is that when you have a successfully deployed protocol that needs new features you revise it.
Once you want to fix these things the RSS road map suggests you ought to do it under a different name. Then there's the issue of syndication culture - apparently civil discourse is hard to have in the RSS community. The culture had become sufficiently toxic that it necessitated starting something new.
Sometime in 2003 Sam Ruby started a wiki to discuss this new syndication format. the email archive on the discussion on the format took 17,944 email messages to decide.
How could this been so hard? You can't imagine how insane the debate on publishing dates was. Feed aggregation was another set of issues.
In July of 2005 RFC 1287 was approved.
The publishing protocol - had it's own set of issues - took from Sept 2004 to March 2006 to get to a stable draft.
Was this a good idea?
Up till now all the standards were created by one guy or a few in a lightweight process.
The IETF process - all done by email/wiki - consensus decreed by chair; may be appealed. Trolls can be banned for 30 days; may appeal. General review by whole IETF. It's irritating and slow, and it takes time. Having said that, Tim's prepared to argue that it's a good thing. The whole notion of a civilized society is that you don't take out disagreements by force, but you go to court instead - it's not pleasant, but it's better than violence. The IETF process insures that nobody can say that the process wasn't followed, or that people didn't have a chance to have a say. The IETF review
Atom is a lot liek RSS2, except -
Feeds & entries must have unique IDs, timestamps, & human-readable labels. Text can provided plain, in HTML, or XHTML, with clear signaling. Can provide both summary and full content. Namespaces & extensibility are clean.
Atom seems sufficiently extensible that there shouldn't need to do an Atom 2.0.
Atom is the general-purpose collection idiom that XML has never before had. A whole lot of business processes require exchanging collections of various kinds.
In Atom Publishing you put something out on the server, and then the server comes back and gives the client a URI where it put it.
Atom publishing should allow for a proliferation of blog authoring software - the APP is the missing infrastructure link in making the WEb writable by everyone.
Most feed-reading applications and libraries handle Atom 1.0 well, except Bloglines.
Google uses Atom 1.0 in its AJAX APIs.
Will Atom be mainstream, or a footnote? The odds are pretty good that it becomes mainstream.
The logical conclusion of web services - the sea of knobs and buttons - APIs being extended like cilia from large databases and lots of ways to interact with ambient data. Right now we're barely squiggling along with the browser. As usual what we thought was the top of the mountain was the first plateau.
all that is mashable recombines - every API and RESTian web service can be combined with every other - what does that really mean? What does it mean that all these things are flailing along out there - how do we make personal sense of this, frame it up? As all these things are spreading and proliferating, we need to be able to control the experience towards us.
Culture emerges on objects - what is the story? The people that are involved in the construction of these services constructed a conceptual model - every couple of years there's a new fad or trend because it addresses some perceived problems with the old construct. These models actually are culture. We have collections of functions and we find different ways to reframe these - that expresses as tribe membership. We need to recognize that that is a playful process.
Collaborative Play transforms culture.
Wanted to build a platform that allows people to share mediated experiences. Started with abstract notions of how to create virtual communities and mediating extremely flexible space of shared experience online. The solution of how to deal with the models clashing is to make it playful - turn the conceptual banter into play. A living social organism that communicates with itself and creates innovation.
A history of Fabric
In a pre-web world there was the MUD, which then became the MOO. The MUD was a way to play D&D online, but the MOO transformed it to the idea that multiple people could become programmers who could create narrative objects. The differences between the user and the programmer were breaking down. You could begin to intuit what it is to participate in an object-oriented programming environment by extending existing objects. The critical planned piece that was never realized was a project called Fabric, which was to integrate the MOO kind of functionality into the operating system - access to networks, file systems, etc in a narrative fashion. At this point the windowing paradigm wasn't set in stone, so the idea of a more flexible paradigm seemed more possible.
Matt shows a moo that treats urls ond web pages and feeds as resources within the game space. You can grep for specific kinds of data in specific tags on a web page.
Verbs have small amounts of Python code that do things, and can be combined.
As Yoz says in the back-channel, "He's exploring the metadata and formats of the web through a text adventure interface."
It's an environment for mashups, combining small verbs.
Matt notes that navigating the text environment with commands like "north" or "east" is difficult when looking backwards at the projection screen.
He shows an alternative view of the space by controlling it as a card game.
There's a shared python interactive interpreter within the playsh environment.
We should use our powerful computers to make the APIs available to those of us who want to be mechanics. The APIs allow hiding of differences. The thickening of the verb layer hides everything underneath. Shows a little of Animal Crossing for the Nintendo DS. Compares this to telephone banking - there's no way of learning from what they operator is doing.
What if we could do all this stuff without xml parsing and learning apis, etc - that's what playsh is about.
Flow calcifies into artifacts. Able to play with the experience as you have it. You can start off the system in total play mode with multiple people, and as that progresses things like tool.. s will fall out of the experience. People in technology are addicted to that moment of play, interacting socially around not-yet-defined experiences.
flickr is an example of a service that grew out of building a game.
Object oriented?
How can we have things that are the same things but seen differently.
player carries around pattern-binding machines which get acturalized at the time of use. "the real is not one thing but gradients of resistance".
Playground Foundation in Amsterdam - multiple threads of investigation of the transfer of playful notions into tasks. How does the literacy in gaming affect the way we'll be able to design computing in the future?
Download available at paysh.org
Amy Jo Kim is talking about applying game mechanics to functional software.
She's trained in Psychology, Neuroscience and CS. She specializes in social games, networked communities, and mobile services.
The talk ais about using game mechanics to creat apps and services that are fun, compelling and addictive.
What is a game? a structure experience with rules & goals that's fun - when push comes to shove fun wins.
5 Game mechanics you can use
1. Collections - status and competition in competing the collection
e.g. Pokemon card collections
Completing the collection becomes a powerful motivator
2. Points
Social Points - e.g. amazon ratings, etc.
once you have points you can have Leader Boards - who has the most points.
leader boards drive player behavior. some systems had leaderboards and took them away - slashdot karma - encourage people to game the system.
Once you have points you can have levels - very very motivating.
e.g. bejeweled game levels, or ebay star levels of colored stars.
Levels punctuate the game expereience - e.g. reaching level 60 in WOW, or earning a new belt in karate.
Levels unlock new powers & access. Ebay has levels of power sellers.
3. Feedbacked -
Feedback draws attention through movement & change. e.g. feedback in Bejeweled. MySpace Mobile giv es you lots of feedback.
Feedback accelerates mastery - e.g.Karaoke Revolution or Brain Training (huge hit on DS in Japan) - math problems and color matching. Gives you both immediate feedback and long-term feedback.
Feedback makes an experience more fun & compelling - that's why Ajax apps are succesful - they feel more compelling become they give you immediate feedback.
Feedback makes mundane tasks more fun - Cooking Mama for DS - teaches you how to cook.
BIMactive - feedback on you physical activity - runner's training tool.
4. Exchanges -
Exchanges are structure social interactions - giving gifts, etc. A basic, primal form of social interactions.
Social exchanges can be explicit (e.g. taking turns in chess), or implicit (i.e. emergent) - eBay feedback has evolved into a tif-for-tat social game. Not built into the system, but evolved on top as a metagame. Trading is an explicit Social Exchange (e.g. trading in WOW or in MogiMogi).
"gifting" is an implicit social exchange - very powerful for driving behavior. Helios - a just announced mobile service targeting myspace generation. MySpace has both explicit and implicit exchanges - add friend is explicit, comments are implicit.
5. Customization -
Customization increases investment of the user.
... and creates barriers to exit.
Automatic customization is fun and engaging - e.g. Amazon customizing your home page based on your buying history. Flickr does this too.
Character customization is especially powerful - not just in games - e.g. profile customization of MySpace.
Looking ahead - expect to see more serious apps that feel like games, and more games that teach real-world skills.
Slides are available at shufflebrain.com
Enterprise Customer Pain Points
As Internet continues to evolve, it continues to expand enterprise partnerships - it's much easier to link businesses together than it was 10 years ago. This becomes integration work for the IT shop - it takes about 6 months for the IT folks to do an integration (that's down from 9-12 months a few years ago) - but the speed of change of business relationships is quicker than that. From the standpoint of businesses today, there's more information externally than internally - need to factor in external information into internal dashboards, etc.
There are a whole bunch of apps not being written today because they're not affordable.
Web 2.0 - rethinking application assumptions -
Think about how middleware is defined - if you assume business organizations and relationships are continuously changing - scalable and long term are not the most important criterai - solution needs are situational.
Line of Business teams have just enough IT savvy to create their own services/solutions that drive their part of the business (Igniting the Phoenix: A New Vision for IT/Sapir)
Applications are disposable in many cases.
"Situational" Apps - What does this mean for enterprises, is there a market for these lines of things? Scripting - people who can put together things in very short terms. "Enterprise Mashups". Lots of customers from IBM interested in these sorts of things - instant dashboards built by domain knowledge folks - can't wait for IT to do it.
Is there an infrastructure that can be built to support the creation of 5 minute applications?
QED WIki - a Wiki but also an environment for integrating data from multiple sources.
Shows a demo simulating a small business with stores across the US. Pulls together store addresses, inventories, maps, and weather data in a quick integration that was done in five minutes.
Linda Stone was formerly at both Microsoft and Apple.
Continuous partial attention - she coined the phrase in 1998. Something that is happening and getting refined to a higher art over the last 20 years - it's an adaptive behavior and we are on the way to adapting right beyond it.
continuous partial attention is defferentiated from multitasking by its impetus - multitasking is getting as many things done at one time as possible. CPA is motivated by a desire to be a live node on the network - we want to connect, to feel important, scan for the best opportunities at any given moment. There used to be a time when people wouldn't pick up a phone during lunch, but those times are gone.
In the golden time of CPA, it wasn't uncommon for people to go into a meeting at spend most of the time doing IM and email - as if we expected our personal bandwidth to increase with the stimuli.
CPA fits into a larger context and set of patterns - theory: we operate with a collective sense of attention. Consider the years 1965 - 1985 - they were about individual expression - it's all about me. We were all about achieving our full potential. We multi-tasked to increase our productivity. From 65-85 the collective ideal was to value self-expression above all else. But being a species that's good at taking things to extremes, we did that and found ourselves yearning for what's missing - a connection to others.
1985-2005 is the era of connecting - the network is the center of gravity - we trust the network and collective intelligence. Playdates replaced the dancing and violin lessons for kids. Enhancing our network and sense of connections. In the early days of Friendster the barroom boast was "I have 3,000 friends." We were everywhere except where we actually were.
Now the 24/7 thing isn't feeling so good - people want strategies to cope. The always on era has created an artificial sense of constant crisis. Is everything really such an emergency? Our ways of using technology would have us believe it is. We're feeling overwhelmed, overstimulated, and stressed.
There are new desires as a result of the start of the pendulum swing, which will create new opportunities. Always-on doesn't respect the fact that there should be cycles. Take email - how effective is it for decision making and crisis management? Not! After all these years of using it we still use it. Wikis are better for brainstorming, IM is better for making a plan. Conflict resolution best done synchronously with high bandwidth, crisis management best done synchronously at any bandwidth. Is it time for some guidelines on how to use these technologies? She thinks email is an attention chipper-shredder (think Fargo).
We're yearning for protection - we've gone from an era of creating opportunity to scanning opportunity, and now are moving into discerning opportunity. Meaning, belonging, protection, and trust will increasingly be what we seek and resonate with.
For the last two decades ease of use has been the mantra of technology - but it's no longer good enough. The new mantra will be improvement of quality of life - does it help protect and filter?
The new opportunity is to move from being knowledge workers to be wisdom and understanding workers.
Great talk!
Attention Economics - A 15 minute course
The first question - what is attention? Are the rules of the attention economy similar to other economies?
Attention is fundamentally about time - while David could talk for hours about this, he really only has 15 minutes.
Attention is time directed to a purpose by people. In a captialist system we talk about scarcity of resources, space, money, capital. An economic system isn't only defined by what is scarce, but it's a good tool to use to determine tradeoffs.
What's scarce in the attention economy, and how does it express itself on the net? CPU cycles are not scarce, nor is storage. Network bandwidth is not really scarce, though people in the US might think so. Money is plentiful, at least for those in advanced economies. Time is scarce - there's nothing we can do to extend the amount of time in a day, lifespans don't increase very quickly. Information is not scarce in this new world of abundance, as a matter of fact we all feel information overload. If you combine the relatively finite set of people with the finite amount of time, people only have a certain amount of time to have meaningful social relationships with a limited number of people.
The scarce resources are time and people.
You can't hoard time- use it or lose it - a perishable resource, like fruits and vegetables, or hotel rooms.
Aggregate attention artifacts - whatever we can do to capture the artifacts of how we spend our time should be valuable - look at how google and yahoo value their clickstream data. It's implicit metadata
Productivity - one of the core insights from Technorati was "how do I make it easy to find out what people are saying about me, right now?". Technorati helps save time.
How can this be applied? Incorporate an understanding of TIME and PEOPLE into your design of your applications. Look at memeorandum - builds fundamentally an understanding of time (when something was created or posted) and linking behavior, to show people what's popular at the moment.Or in Technorati, you can use your friends as a social filter.
Attention is both a currency and a perishable - make it easy to create and express attention.
Hyperlinks are votes of attention
Create opportunities for economies of scale.
Runs on about 2000 CPUs in SF. Second Life is a virtual world, about 4000 people online at the moment.
As the pain of creation and participation gets harder, participation goes down - but in Second Life the pain is high, but lots of people make things from scratch, building objects or textures.
35% of adults say they spend more time online than working.
Second Life is not a game - no winning or paths or anything. Everything is built by the residcents in second life. Not a subscription model - you pay for permanence. 240k items were bought and sold in the last 30 days. The average transaction price was > $1 US.
Unanticipated consequences - some users created alien abductions, and they'd then give you a t-shirt that said "I've been abducted and all I got was this lousy t-shirt".
They have about 90k hours of use per day. They're growing at about 15% per month.
About 15% of users write code, in a c-like language with lists - not an easy environment.
UC Davis - virtual hallucination project, to simulate the symptoms of schizophrenia - created by one resident in two months.
Use is gender-neutral, median age is 36.
Virtual worlds seem to embody good learning practices - there are hundreds of classes a week in how to use Second Life.
There's lots of legitimate peripheral participation as a way to learn.
There are 17 universities teaching classes in second life.
Party 906 10oth Ave 9 - 11 tonight for those interested in being recruited.
Root is an "attention exchange"
Web 2.0 = Attention 1.0 (huh?)
Is attention about money, or is it about time? He's from NY, so he focuses on the money part of attention.
Information attracts attention. Receiving attention makes you influential.
Attention macroeconomics - why attention now? Because web services have enabled the recording andsharing of attention choices in real time (a la last.fm).
Root is an open exchange for the attention economy (huh?). The back channel expresses confusion about what this talk is about.
He's now talking about something called promise to pay attention, which leads to attention bonds. They think those attention bonds can be pooled and traded.
By sharing your online activities (who are you emailing, what are you reading online, etc) these impressions become worth money.
They've founded something called the Attention Trust, a non-profit to work on these ideas.
They have something called a Root Vault - you can send your clickstream to your Root Vault, and then actually see what you've been paying attention to.
This one left me sort of confused - but it seems like there's something worth thinking about here maybe.
Felix starts out talking about MyWare, which sounds like Spyware, but is basically people spying on themselves - if people tell last.fm what music they're listening to all the time, last.fm can infer other music they might like and make recommendations, based on the "knowledge of the crowd".
They have some plugins for audio players that submit what tracks you listen to to last.fm - they currently get 8 million submissions a day. That builds up a catalog. There is a lot of noise, but currently 8 (out of 25) million tracks in the catalog have clean metadata.
They can calculate user-to-user similarity (who else likes what I like, and what else do they like?) and artist-to-artist similarity (what other artists might I like).
The backchannel chatter has it that you have to let last.fm record what you listen to for about two months before it really works well for you.
Dick is giving his highly anticipated followup to his identity 2.0 talk from last year, this time entitled "Who Is The Dick On My Site?"
Dick is talking, in a highly entertaining way, about identity issues in the online world.
After briefly outlining the problem he talks about federated identity (Liberty and Shib) and delegated identity (like SalesForce.com), then talking about Yadis (a lightweight identity model which doesn't require explicity trusts between identity provider and service provider). Now he's talking about what he calls Pull models - including MS infocards, and SXIP. They're hosting a Digital Identity Exchange BOF at IETF in Dallas.
Ning used SXIP to create a user registration app in twelve minutes.
SKIP 2.0 works with existing browsers. Doesn't require trust between IdP and SvP . Has some nice UI for user controlling granular release of personal data to service providers. There's also the implementation of something called the authoritative site, which vouches for good info on the user (like verified email address). If I understood it right, there does need to be explicit trust between the service provider and the authoritative site.
Last night Rael asked the audience "who cares more about the content in their RSS aggregator than their email inbox?" - only a few raised their hands, but other presenters are asking the question or remixing it into their presentations.
Perhaps an emerging thought string.
Felipe is the VP for Software Development at Amazon web services.
They are trying to enable humans into the software process. Normally you think of AI as being human asking something of the computer - they've turned it around into something they're calling "artifical artifical intelligence". They're providing a Web services API for computers to integrate this by allowing programs to make requests of humans. People can get paid for this work. This is what someone in the backchannel last night called the sweatshop of the future.
I didn't catch the name of this guy, but he's demoing a very responsive multipoint touch for very intuitive physical interaction with objects on the screen.
He shows a photographer's lightbox app where you can drag photos around with one finger, and then with two fingers can zoom in and out on the screen.
Then he shows a video app where he's running all 36 channels of Time-Warner cable in windows simultaneously.
He shows a mapping interface that allows for easy zooming and rotating of map data from.
THe photo app features a resizable virtual keyboard.
He gets the loudest round of applause yet heard at the conference.
Ray Ozzie, now from Microsoft, gives the opening keynote of the morning.
RSS as the DNA of the future Web
Enables the weaving together of composite apps, much like people do on
unix with piping, or with GUUI apps using the clipboard.
But the Web is a bunch of silo'ed sites. We've done work with interchangeable
data formats, but where is the clipboard of the Web?
Why isn't the clipboard for the Web the clipboard?
Demonstrates a new feature, Live Clipboard, which allows cutting and pasting of structured web data - he demonstrates copying calendar data from Eventful and pasting it into Windows Live Mail calendar - bridging data from one place to another. He shows this using Firefox instead of IE - I guess to make the point that it's not an IE thing.
Now he demos putting feed data into the ckipboard. Clicks on a clip icon on a blog and pastes the feed url into Bloglines.
Next he shows clipping his current location data from his MSN Spaces profile (which it gets from his Wifi enabled Windows Mobile phone) into his Facebook profile - then he subscribes Facebook's profile to an RSS feed of that data from MSN Spaces. Then he feeds that data into a map service, and subscribes the map service to the location feed of several folks.
Then he pastes that data into Excel, where it displays the structured location data in cells.
Finally he shows clipping thumbnail photo data from Flickr and pasting directly into Windows photo viewer, where it gets not only the thumbnail but the whole photo.
Ray asks for the community to get involved, and promises details will be posted on his blog.
Technorati Tags: etech06
When I went into lunch this today, I saw a fellow with a bass clef dyed into the hair on the back of his head - so of course I had to sit down and chat with him.
Turns out he's not a bass player, but he is Robert Kay, the brains behind MusicBrainz. MusicBrainz is working on becoming a public source of the kind of musical metadata like is commonly found in the commercial Gracenote CDDB service, or (as the web site says), a "community music metadatabase."
Robert told me that he got pissed off when, after he had, as a volunteer, input data on hundreds of CDs that CDDB didn't have, they went commercial and claimed the data as private property. He decided to do something about it and create an alternative.
They've currently got over 225,000 artists and 4.5 million tracks indexed, and more coming in all the time. They've also got some cool software that compares audio signatures and matches music in your personal library against tracks in the MusicBrainz database and adds the proper tagging to them.
Very cool, even if I didn't get to meet another bass player.
I'm in San Diego this week for O'Reilly's Emerging Technology Concerence. This morning I'm taking Kathy Sierra's Creating Passionate Users 2.0 tutorial.
The first thing Kathy says is that we'll be talking about passion - not people who like your products, or use products, but the kind of passion that they bring to hobbies like skiing, music, their pets, etc.
Reverse engineering passionate behavior - People with a passion... learn, show off, connect, continuously improve, spend time, elevate the meaning, evangelize, spend money.
Look at attributes of passion, and see if we can encourage passionate behavior by working on those.
The one thing we find where there is passion, there is a user kicking ass. Nobody is passionate about something they suck at. That requires constant learning, growth, and progression - people are not passionate about something that has no potential for growth. Passion requires learning and improvement. Learning increases the resolution of the experience. How can we make that investment of time worth it?
Need to think about what that thing is that we're going to help users get better and better at. Many companies only help people find out about the company and its products, but don't help the users actually get better at doing something.
Nobody will stay passionate about something where there isn't more to learn. Kathy asks how well you're doing know at helping your users kick ass at what it is they care about.
There are things that the brain cares about - we can use those things to get our foot in the door to get someone's attention.
The brain has a built-in crap filter - there's a lot of new research on this, coming out medical research in Alzheimer's (like Eric Kandell) . What does the brain care about? Weird stuff, that's new or novel - the brain is constantly looking for expectations to be met, and it's jogged by that which doesn't meet the expectation. The brain also cares about sex, and beauty, and innocent and cute, and having fun (all mammals have a high play drive, because it's how we learn to survive) - don't underestimate the power of fun. We tend to suck all the fun out of technical things - why? There are simple tricks to use to keep people's brains awake. Leaving things unresolved can get the brain's attention - advertisers use this a lot to get your attention.
Things the brain cares about: Unexpected/Novel, Scary, Sexy, Beautiful, Innocent (young, cute), Funny, Faces, Unresolved.
Conversation beats formal lecture!
Conversational language in any form of documentation makes the brain think it's in a conversation and it has to hold up its end. There's lots of research to support that conversational tone improves involvement and subsequent recall. Thre's some study of this in The Media Equation.
We want to talk to the BRAIN, not the mind. They don't operate on the same goals.
Now that we have their attention we need to keep them involved and engaged. We have to get past the "suck threshold" )how soon do they stop hating it?) - the longer that takes the more chance you have for attrition. But some products have a long term out into the future of progression, but takes longer to get pas the suck threshold.
Why does anybody snowboard twice? There's a clear picture of what it looks like when you get up that curve. In your product or service, have you painted that picture? And then, make sure there's a path to get there. It can't just be "here's the end state, figure it out".
You need - A way to recognize expertise, a meaningful benefit, and a clear path to get there.
Why? Who cares? So what? The technically accurate reason is often not compelling. "make sure you connect the abc to the xyz" is common, but not compelling. Then they often explain why you need to do that, (who cares?) and at the end when you're just about to kill them, they finally give you the answer about why it's important. Honda calls this the "5 whys" - when you ask "why" five times, you finally get to the heart of the matter - you should give that reason first! One other technique is to get to that final point where you say "because you'll never have sex again", or "because you'll get fired", and then take one step back, and that's the thing you should say first about "why".
Need to think about imparting understanding, not facts. The crap filter controls long term memory - so invoking emotion or providing motivation will make the brain more willing to actually store information.
Neurons must fire! People learn from mistakes. Can you provide users some experiences where things go wrong without pissing them off?
The Smackdown Learning Model - Offer two conflicting but compelling viewpoints - people will think about the unresolved part of this.
Just in Time vs. Just in Case - the "oh crap, oh cool" method. We want users to feel the need for something ("oh crap") and then offer the solution ("oh cool"). It's not about the answers, it's about the questions.
We have to keep users engaged - Flow, the Psychology of Optimal Experience They way to think about when you're in "flow" - e.g. while you're programming you keep thinking you're just one compile away - then you look up and seven hours have passed. You're so engaged and focused that you don't even know that time is passing. You get into that state because Knowledge and Skill are in balance with perceived, meaningful Challenge. Game developers have this down. And it's important that you know it's important to solve the problem. If the challenge is too hard they'll perceive it as not worth it and drop out. If the challenge is too easy, people won't care and they'll drop out.
Another interesting book: "Don't Make Me Think : A Common Sense Approach to Web Usability (2nd Edition)" (Steve Krug)
Need to think about how to help people stay in character - if they have to break character to think about the interface, for example, they've lost the flow.
What do game developers know about putting people into flow? The spiral experience model - build interestt, provide motivation, then give a payoff. What do you do during the payoff phase? The longer you take to get to the payoff, the better the payoff needs to be. Games have the concept of the "next level" - spend some t