Report from Mashup Camp

Mashup Camp Breakfast

Originally uploaded by sogrady.

Before I get to all of the good stuff from today, I have one complaint: start time. I haven’t spoken to David Berlind about it yet, so it may well be that they had a good reason for beginning the day at 7:00 AM, but that was just brutal (particularly since I didn’t get to the hotel until about 1 AM). That would be early for virtually any conference, but as Christopher says and Alex and I grumbled this morning, I don’t know any developers that are up at 7, let alone working. If they’re looking for advice with respect to making the Mashup Camp sequel better, my feedback would be: start later. Much later. As the Beastie Boys say, sleeping late is much easier on your constitution.

Apart from the unholy start time, however, this has been a phenomenal first day. We’ll see what tomorrow brings, but this is shaping up as the best conference I’ve attended in a long, long time. Maybe ever. While that undoubtedly smacks of hyperbole, any conference that includes the following:

1. A session with a room full of developers talking mashup economics
2. A session in which Chicagocrime.org is deconstructed piece by piece
3. Professor Lessig soliciting opinions on how the Creative Commons can help create a Mashup ecosystem
4. And a free ranging discussion of the possibilities of search, open, vertical or otherwise

is pretty much worth its money. All the more so since there was no entrance fee. That, of course, doesn’t include the hallway conversations with and introductions to folks like Wayfaring (and CNET)’s Mike Tatum, Technorati’s (for now) Niall Kennedy, Partysync’s Anu Nigam, and Sharpcast’s Adam Tow (of Apple Newton fame).

The ratio of signal (read: developer) to noise (read: everyone else) here is outstanding; the best of any conference that I’m aware of. To be sure, there were a few obvious marketing types doing intros this morning, but the overall quality of the people here is outstanding – despite the inability of some highly qualified folks like Alan Taylor to attend.

Because it’s late, I won’t publish full notes on each session, but here are a couple of key takeaways from the ones I was able to make (it was really difficult to pick):

• Business Models in Mashups:
  This session, moderated by Dave Nielsen (StrikeIron) and Eleanor Kruszewski (Yahoo – and Eleanor, if you run across this, we’d love to set up a briefing about the Yahoo Developer Network), was outstanding. The least technical of any of the sessions I attended, it was a really free ranging and engaging discussion of what some of the potential economic models were for monetizing the effort put into sites (build a viable business, build a business to flip, build a portfolio to get hired, etc), and what some of the factors were that factored into success or failure here. Interesting datapoints: asked about who was creating Mashups and on what basis, 12 were doing it on behalf of their employers (i.e. DeWitt for A9), 20 were attempting to build their own businesses, and 5 were doing it purely on a hobbyist basis. Irrespective of the group, there was a general unenthusiasm for the prospect of taking on VC money at the expense of control.

• Chicagocrime.org, Scraping and Reverse Engineering:
  This was probably my favorite session of the day, not least because I got to meet one of the developers I personally invited to Mashup Camp on behalf of the organizers, Adrian Holovaty. If you’re not familiar with Adrian, you should be: he’s one of the Python devs behind the popular DJango framework, just created a Python library for those condemned to using Exchange on a non-Microsoft platform (such as yours truly), leveraged his success on the web into a gig with the Washington Post, and, of course, is the creator of the brilliant Chicagocrime.org. I’d break it down for you as Adrian did, but I couldn’t do it justice. A couple of interesting tidbits about the site that the Chicago PD found ‘interesting’ – it runs Postgres, leverages PostGIS for some of the computational geographical calculations, relies on the geocoding (address to lat/lon) API from Yahoo, and was reverse engineered using the Live HTTP Headers extension for Firefox. Adrian’s key lessons in doing exactly what the original data source – the Citizen ICAN site – was unable to do?
  1. Make the site browsable by every data type (beat, neighborhood, ward, street, type of crime, etc)
  2. Everything that can be a link, should be a link
  3. Expose searches and filters as a permalink

  If you get the chance to hear Adrian speak, I highly suggest it. This was brilliant.

• Creative Commons and Mashups:
  Before last week, I’d never actually seen Lessig speak in person. I’ve caught many of this talks before either on webcast or through things like IT Conversations, and I’ve enjoyed his blog, but I never got the chance to see his particular style – the Lessig Method – live and in the flesh. Having had that opportunity at OSBC, I can safely say that Lessig is one of maybe 2 or 3 people in this industry that I would pay to see speak on a regular basis. He’s so good that you do worry a bit that the style can impress so much that the substance gets lost, but he’s definitely worth the price of admission. Today was very different, in that it was the sort of scripted, seamless session that he put on last week but rather a more back and forth discussion. My first question was whether or not he believed that the Creative Commons ran the risk of having its “Some Rights Reserved” message lost in a “they’re all about free” meme (as I heard this morning). His reply was “Absolutely. There is every risk of that happening.” He acknowledged that in the grand scheme of things, the Creative Commons are not professional marketers and they need help getting the word out. My last question, was whether or not he believed that the Creative Commons could play a role in simplifying the issues around fair use, liability and so forth that plague Mashup developers. His response was basically not really, because the laws in the space are so convoluted. His plea was that we all a.) begin to catalog a set of normative or reasonable behaviors in the space to form precedent, and b.) to argue more strenuously for the reform of said twisted legislation. I’m on board, and of course we at RedMonk – having licensed virtually every piece of content we have under CC licenses – are happy to do what we can to help.

• Mashing Up Search:
  Last, but certainly not least, was a session led by A9’s DeWitt Clinton that discussed the possibilities of search. The interesting thing, as I was discussing with several folks at cocktail hour, is how so many have prematurely (IMO) concluded that search is over, and that everything that can be done in the space has been done. Nothing, as near as I can determine, could be further from the truth. Topics of discussion included vertical search (does a query for Vioxx return litigation attorneys or relevant drug information?), APIs for search, Yubnub’s community oriented search, and – of course – my personal hobby horse, the latency of search results. It would have been nice if Google and Yahoo had more folks here, but DeWitt did a fine job of articulating some of the challenges and opportunities in front of A9’s search efforts.

If tomorrow’s half as good as today – and it should be, because we’re expecting to see lots of demos and the like – I’ll be a very happy camper. Meanwhile, kudos to Berlind, Doug Gold, Mary Hodder, Kaliya Hamlin and the other organizers for doing a bang up job. If we can get the start time problem fixed, this will be a near perfect conference.