Mashup Camp brought together some 400 people interested in developing web applications quickly by combining mutiple data sources or services in interesting ways. This being my first "unconference", I found the just-in-time and minimal organization of the camp an interesting study in the balance of forces between anarchy and social grace.
The camp itself ran Wednesday and Thursday. Mashup University, a more traditionally structured single track of sessions, ran the two days prior to a handful (60?) of camp registrants who signed up for the university as well. Mashup University provided Mashup Camp sponsors with a soapbox (session slot) to tout their wares, as well as much needed funding to cover the costs of the food and venue for the Mashup Camp itself.
As you can see from the schedule, the sessions were roughly split between major hitters such as Adobe, AOL, Intel, and Microsoft, up-and-comers such as Plaxo, and a variety of very early startups. At least one of the startups opened their session with "We don't have a business plan; we're hoping you can tell us what you want us to build." Sounds like money looking for a pit.
Google and Yahoo were conspicuously absent on the University days, and very light on the camp days. What's up with that? This isn't just a crufty Microsoftie taking pot shots - several of the attendees wondered aloud why/if Google had abandoned them.
Who Do I Trust?
A recurring theme throughout the mashup event were questions like "What data/service providers can we trust to support the services we're building our business on?" and variations such as "What happens two years from now when they pull the plug on the service? Am I SOL or do I have some sort of recourse for ongoing support?" and even "What if my mashup really does become popular, and profitable? Are they going to pull the plug on me because of load or competitive issues? Or just move into my space?"
The answer to all of these is simple business: get to know your service provider(s). Look at their track records for support, their policies, and their attitudes. If you're building a business around assets (core services) provided by someone else, common horsesense says you'd better do everything you can to protect your access to those mission-critical assets. Demand service level agreements (SLAs) that include not only network uptime minimums but also termination and transition terms with real monetary teeth if the provider lets you down. Novices may balk at the idea of paying for SLAs for a "free" web service, but in the real world of business the smart money hedges against a greater loss. Spend a little now as insurance against losing a lot later. Heavy SLA penalties for failure to perform also give the provider serious motivation to pay attention to operations and accountability rather than the easy money of selling services ahead of actual capacity.
One thing is absolutely certain: Do not accept anyone's promises or goodwill. Get it in writing. If you accept anything less than contractual commitments (and their associated costs), then you can't really say you're in business; you can only say that you're busy.
Because the mix of participants at Mashup Camp was highly tech-entreprenurial, there were a lot of folks there with a lot more technical ideas than business experience. Questions and passionate debates about whether a service can be trusted to be around in 2 or 3 years just don't come up that often at more line-of-business events such as TechEd.
Caught in Passing
Various folks I bumped into at Mashup Camp:
Michael Scherotter of MindJet. "the MindMap guy"
The Dara-Abrams clan (Alec, Benay, Cassie, and Drew) churned up quite a bit of discussion. Very interesting people. Teen Cassie led two Mashup Camp discussion on the topic of addressing the younger generation of web users, but spend most of the time fielding questions from those pesky older folks. Officially, Alec is with Sony, though it might be more accurate to say Sony is with Dara-Abrams.
Daniel Charles of Digital Ketchup. "The Software Condiment Company" Finding a catchy name is half the battle of breaking into the market. Plus, they're in Toronto, Bruce's neck of the woods.
Cameron Jones, PhD candidate at University of Illinois at Urbana-Champaign. Cameron came to Mashup Camp to study the mashers as much as the mashups they produce. He's digging into questions like what the development patterns are in mashups, and whether those may be beneficial in other collaborative contexts such as academic research.
Krishnan Iyer asked about Vista application compatibility and whether the app compat info available on Microsoft web sites would be provided in a digestible (RSS) form anytime soon.
Johannes Ernst of NetMesh led the session on managing identity across multiple vendors mentioned in the previous post.
John Planzer of AOL was also in the identity session. We concurred on a number of points during the discussion, and talked briefly after the session as well.
Bartosz (Bart) Solowiej came out to Mashup Camp from Washington DC to demo his traincheck.com mashup app that provides next train times for a given station via SMS or email. He planned ahead - added the San Francisco BART and CalTrain timetables to his app to make it relevant to the folks here at Mashup Camp. I pointed him toward the Windows Live Messenger bots so that he could handle IM requests as well as SMS.
Eric Keosky-Smith of Incendo Marketing was demoing a small-community portal system whose business model is to address the lack of marketing focused on the middle of the U.S. "We stay away from the coasts and the metropolitain areas, and instead focus on a thousand communities of a thousand consumers." Interesting premise.
Joseph Smarr of Plaxo was there showing ways to tap into address books across domain boundaries. He found a way to talk to just about everybody at the event, but I think Scherotter still has him beat. Scherotter runs circles around everbody. Even those who saw the MindMap pitch at TechEd only a few weeks earlier! ;>
StrikeIron and Mashery, both service providers targeting mashup developers, had good representation at the event, though Mashery is still very much in the formative stages at this point. We'll see how that forms up over the next few weeks. StrikeIron's model of aggregating other services under one roof bears some looking into.
Max Dunn, self-described as a "retired serial entrepreneur", led a couple of Ruby on Rails discussions. When asked if he answered to "rabid Ruby on Rails advocate" he replied "Well, people say I get excited when I talk about Ruby on Rails!" Yep, that qualifies.
For Microsoft, we had Ken Levy talking about Windows Live Messenger APIs, Steve Milroy on everything Virtual Earth, Scott Isaacs on Windows Live gadgets, Trevin Chow fielding questions on Windows LiveID from the back of the room, and Todd Biggs doing the sponsor thang.
For the Next Mashup Camp
Two suggestions for the next Mashup Camp:
1. Figure out some way for people to submit session topics online, to eliminate the ridiculous paper chase scheduling system. Everybody in attendence had a minimum of two web connected devices on their person. Wiki doesn't handle mass submissions well; find a queue front end and let people submit into the queue, and chew that up into wiki pages on the back end.
2. Host a panel discussion with a group of teens to talk about web use amongst the younger generation of Internet consumers. Then perhaps Cassie can actually have a discussion with developers about her questions instead being hounded by their questions of her. Cassie's bright, technical, and a teen, but a sample size of one is not a good survey to base mashup business upon.
Over all, Mashup Camp was a fun, free event for developers. I look forward to the next one!