Monthly Archives March 2009

I love it when a plan comes together…

So Simon Willison (@simonw), one of the architects of The Guardian’s Open Platform and co-creator of a modestly popular web frameworks is here at SxSW and gave an informal talk (on Zeppelins, of course – what else?). Freebase community manager Kirrily Robert (@skud) saw my tweet and proposed a meetup. After iteratively solving the three body problem, we put out the word on Sunday morning for a meetup on Sunday evening… SemWebAustin @juansequeda and Freebase @jameshome each pinged their 1-neighborhood and next thing you know I’m sitting next to Jure Cuhalev of Zemanta and machine learning machine @Nikete trying to orchestrate overflow seating for 25+ data geeks.

The reason for the gossip-column style of this post is to show the size and breadth of the data geek crowd. James Home and I agree that we need to turn out this Cyrus’ army of data geeks to take over a much larger part of SxSW next year. We need talks on column-store databases and hadoop, linked data and the construction of the data commons, how NLP and machine learning can power inspiring audience-driven websites, on the developing grammar of Information Visualization, on Processing and Prefuse and R. Pete Skomoroch, Mike Driscoll and Christian Chabot all ended up skipping SxSW this year; we need them leading a panel discussion on how to visualize >10M point datasets with limited-bandwidth desktop and web interfaces. I’d like to hear Deepak Singh and one of the @cloudera’ns drop science about scalable cloud computing.

The evening was just informal mingling and conversation, but on request of request of @mndoci and @dataspora, here is our name-droppy slice of the whirlwind:

@mrflip: Learned about how Zemanta is already putting Linked Data and NLP together to make blogging better. Jure is excited about infinite monkeywrench and might be brave enough to pre-alpha its inchoate HTML munger. Got to hear what Blaine Cook of Osmosoft is doing to solve the fractured twitter/facebook/’ve-never-heard-of ecosystem, and he gave some great feedback on our upcoming Twitter Census. Also got to learn, after pontificating that OAuth is hard, that I was talking to its architect; a great discussion with Blaine and ENTP Uruguay Evan Henshaw-Plath followed about the Rails authorization/identity/authentication stack.

Mike Migurski of @stamen is going to get together with infochimp @dhruvbansal to push the Open Street Maps dataset into Amazon Public Data Sets collection. Harper Reed of Threadless was running off for a 6am (ugh) flight to babysit servers in Chicago by the time we chatted, but pointed towards his Chicago Transit API project. His post on Hidden APIs is a great read BTW. Ran into @Slicehost Matt Tanase at a party after; Rackspace is getting much Cloud-ier, including a 1.5cents/hour pay-as-you-go 256MB slice offering. I’m hoping to talk later about our MachetEC2 project and get his thoughts about how to put open data on tap in the cloud. Jon Pierce and I discussed the Mets’ chances this year and what he sees for big data startup possibilities. Only got to briefly intersect with Andrew Turner about open geocommons, and was chagrined to learn I was shoulder to shoulder with one of gnip but didn’t get to chat. Hope to fix that later.

This meeting alone made SxSW worth it, and I’m looking forward to more discussion later. You can stalk me on twitter as @mrflip or at By the way, I’m giving a lightning talk on Open Data in government at Fiddler’s Hearth, 301 Barton Springs Rd at 12:30 — drop by or catch the webcast later.