I ran over to Togo’s with Andy Smith to pick up the sandwiches, a nice opportunity to talk with him and get a bit more information on Flock. Beau Lebens, who I worked with on the Blogger API PHP code so long ago, came down just before I left and we finally met after so many emails, IMs and phone calls. Taller than I expected, he’s in the blue shirt at the back of this photo Scott Beale took. Much schmoozing with Ho Jon, Beau and Alex Russell, a developer at JotSpot. Alex had some good information on changes at JotSpot’s functionality and pricing. The WhumpMeister himself, Bill Humphries, showed up around then too.
Around 2:20 Chris Messina put up the anticipated Flock demo. Looks good to me. Visually appealing with a nice understanding of how to surface two-way web functionality. Key early features are:
- built-in social bookmarking, including groups and watchlists, which they call breadcrumbs;
- a blog authoring widget that uses ATOM to communicate with pretty much any blog tool; and,
- history search that works against your search history and breadcrumbs.
As part of the blog authoring piece Flock built a component called the Shelf, a scratchpad to which you can drag fragments as you surf and then pull into the rich text editor field. It’s intelligent so if you highlight and drag a bit of hyperlinked text or a photo to the Shelf it gets as much information as possible–the hyperlink as well as the text, for instance, or the Flickr URL for the photo if you grabbed it from that site.
At 3:00 I sat in on an AJAX performance session lead by Jonathan Boutelle. I really didn’t learn much but that’s more from my low level of technical exposure to AJAX. Others probably had a better experience. I was also having physical difficulty sitting on the floor with my legs crossed and LittleSteven on my lap so standing up and walking out felt good.
Waiting for the next session back in the big room where I got a chair at the center table (yay!), Chris did another Flock demo. I got another nugget, that the blog authoring widget can handle any arbitrary number of blogs that you have.
At 4:00 Riana Pfefferkorn lead a session on search engine marketing, mainly for Google where she used to work. Covered some of the history and practice of AdWords, a few common misunderstandings (such as click fraud tactics like trying to blow your competitors budget don’t really work), recent changes and things that do work.
Beau also took a chunk of the session to talk about his experience working with Yahoo’s paid search inclusion. He has a more positive view of it than many other techies might but that’s due to his greater familiarity with the real world implementation rather than coming from a, um, highminded philosophical perspective.
5:00 was Tom Conrad talking about his company Pandora and its Music Genome Project. MGP is the result of having hundreds of professional musicians scoring songs (mainly rock and now getting into Latin) on several hundred parameters, time consuming painful work but the musicians did get paid for their time. Originally they licensed the technology to, for instance, Amazon for use as “if you like this, you’ll probably like that” recommendation engine but that business model wasn’t working.
So now they’re coming out with a proprietary music service. Based on the analytical scores Pandora will let users (once it launches, closed beta is on now) specify one or more artists and/or songs to create personal radio stations. Streaming, not download, and because they’ve set up under the DCMA statutory licensing provision has to follow certain rules. Nice UI for the client app, using OpenLaszlo, and the songs served will frequently surprise listeners by reaching into the Long Tail.
BillSaysThis 