What a fantastic day at the Web 2.0 Conference (and Expo) in San Francisco. It was really fun being surrounded by so many people, most of whom are doing such interesting things using (mostly) such cutting edge techniques and technologies. Today was a very fun and engaging day that covered a wide swath of the Web 2.0 landscape.

My first session was with Jia Shen of RockYou. He talked about how his company approaches Facebook application development. A lot of his points were quite useful. He highlighted the need for rapid development so your product and brand don't get stale. He likes telling the story about how SuperWall was developed in a weekend - started Saturday morning, launched Sunday at 3 am! He highlighted the need for brainstorming and coming up with fun, creative apps that will engage their existing user base (which is surprisingly 60% women) while attracting new Facebook users to sign up. One of his major points was that one's approach to app development should be from an advertising and marketing viewpoint. He promoted A/B testing as a good way to figure out the optimal flow for the install process since that's when the majority of potential users get lost. Overall it was a really fun talk about a very interesting topic.

Next I attended a session that was more philosophy than tutorial but was pretty cool just the same. Joseph Smarr of Plaxo touched on the very important topic of Data Portability. In a nutshell, he's basically concerned with the fact that a user isn't able to define him/herself in one place and have that information show up everywhere. For example, I love The Magnetic Fields. I love them everywhere, all the time. But despite this, I'm sure every site on which I have an account doesn't list them in my favorite artists section. Wouldn't it be neat to define my favorite artists in one spot and have that show up everywhere? Or another example. I'm friends with Jorge. I'm friends with Jorge everywhere. But I'm sure there's a site we're both on but don't realize we're both on and therefore aren't friends (probably eons.com). Why can't I store friends information in one centralized spot? If you stop and think about it, it's actually a really cool idea. I friend you on Facebook. It updates my account. That information gets pushed everywhere else, and if you're anywhere else we're friends there too, automagically! I think this is (or should be) the direction that Google is headed with its OpenSocial platform. But the clear question is, what's in it for the sites? If I can get all my Orkut contacts on MySpace - whose site I absolutely love - why would I ever go to Orkut? This was a question I posed and he had a very interesting response. Imagine I can take my LinkedIn contacts everywhere. Wouldn't I spend that much more time on LinkedIn solidifying my contacts since I knew if I did it there I wouldn't have to do it again anywhere else? Same goes for finding friends taggings on ma.gnolia or tweets on twitter. The logic isn't foolproof - sites that don't have any particular niche become sorta useless - but it's the first real argument I've heard for sharing data. Cool talk.

At 1:30 I sat in on a tea party with 3 panelists who talked about nothing. That was frustrating so I left after 20 minutes and caught the end of what seemed to be a pretty neat talk about JavaScript hijacking. All I really caught was the conclusion which detailed how only a few of the JS Frameworks actually implement protection for it and how we as developers can protect against it server-side. But from what I saw it looked like a really good topic to go more in depth with from both a curiosity and actual security perspective.

The final session of my day was entitles "Design your API" and was presented by two Twitter guys (who assure me they're getting paid). Actually one officially works for Twitter and one was brought in to help build the Twitter API. Twitter delivers an astounding 80% of its traffic through its API. Which means this thing is getting hammered. Which means it needs to be done right. There were several key features of effective APIs that they highlighted including:

• Have a dedicated subdomain for your API and implement caching immediately
• Version the API from the get-go
• Make life easy for Flash developers, i.e. pass back skinny XML and not heavy RSS
• Help your developer community as much as possible and listen to their feedback
• Make it secure
• Make it RESTful
• Make it awesome

Then there were the keynotes. O'Reilly got people all O'riled up and read a 50-line Rilke poem (which was not part of the O'riling). Max Levchin (dude who created PayPal and Slide) basically told us to, in the words of Winston Churchill, "never surrender." I think Churchill was actually talking about Internet startups too when he said that. Can't remember. Amit Mital from Microsoft showcased their newest product "Microsoft Mesh". This new product basically syncs files across any number of devices and greatly eases the ability to share files with others. I feel like there's going to be a fair amount of overhead to learn how to actually use the system for the average user - Microsoft has a pretty spotty history in terms of designing for the computer-savvy. I know, for example, that no member of my family would be able to figure out what he was showing in his video. (Paraphrasing) "You open up the mesh, attach your devices or link to them via your network, open up your target folder, and drag and drop files you want to add to your mesh. Files you want to share can be dragged to your shared mesh, etc etc." My Mom and Dad would be mouth agape after "You open up the mesh"... It's just too complicated. Not to mention it only works with Microsoft products, though they claim Mac support is coming. So we'll see how useful it actually turns out to be. It's an okay idea though. Steve Gillmor likes it: Microsoft Says Yes With Mesh

Clay Shirky's keynote "Here comes everybody" was the most interesting thing I'd heard all conference. He set the stage by taking us back to the early-to-mid 19th century. Streets of London. Gin carts (carts!) are being pushed up and down the streets, the salesmen pedaling their wares to keep an entire generation inebriated and useless. Similar alcoholic binges were occurring en masse, all over the world. Then the human race sobered up and actually started doing stuff. And we get the period in world history known as the Industrial Revolution. Now it's his premise that we are coming out of a similar period of cognitive suppression. And what, pray tell, does he believe to be our drink of choice? Wait for it. Sitcoms. TV! How many hours of TV do we watch as a nation, or worldwide even?? He threw numbers out (which I can't completely remember) but they were in the tens of billions of hours. One overall number was in the trillions. Needless to say, and anyone could've told you this, we watch a lot of TV. But what's happened over the past couple years? People have been turning away from the tube and turning towards the web. Now instead of watching reruns, people are editing Wikipedia articles. Or even if they're not doing something that cognitively engaging, if they're just editing their 1,000th LOLCat, at least they're engaging their brains in some fashion. He told a funny anecdote about his friend's kid watching a movie. At some point she got up and walked behind the television set and started rooting around in the wires. "What are you looking for?" his friend asked. "The mouse," she replied. We now expect our media to be engaging. We are coming out of the decades-long stupor produced largely by mindless TV and are starting to participate in a worldwide movement of creativity and collaboration. Typing it all out, it is pretty exciting to be a part of.

In Conclusion,

The revolution will not be televised...