EuroClojure 2014: Notes & Impressions
2014-06-30 in clojure
Although Clojure isn't our company's primary language, we've decided to go ahead and visit this year's EuroClojure in Krakow — I've always been impressed by the community behind the language, and the list of talks and speakers, as well as the recent developments in the Clojure ecosystem, guaranteed it won't be boring.
Long story short: it was worth every minute spent on travel — plane tickets were ridiculously expensive so we decided to drive. Approx. five hundred miles in one direction does take its toll, so I turned out to be exhausted after the trip, a sentiment which lingered throughout our stay in Krakow. Oh well, win some, lose some — Central Europe is quite nice this time of year.
Luckily for me, I still don't have a driver's license, so the burden of driving fell onto my good friend and business partner, Boni, who took it like a champ that he is. :)
Don't remember the last time I've attended a conf. where I was totally thrilled and jam-packed with new info after every talk. #EuroClojure
— Nikola Plejic (@nikolaplejic) June 26, 2014
The Conference
The conference itself was top-notch, both content-wise and organizationally. I really liked the more relaxed, informal, hacker-ish atmosphere. The venue made that pretty easy — a beautiful, cozy museum of Japanese art and technology, right next to the southern bank of the Vistula river, not too far from the city center.
As I tweeted at some point, I don't remember the last time I went to a conference where I thoroughly enjoyed each and every talk, and learned a bunch of cool stuff along the way. The eclecticism of the talks was impressive, and I applaud all of the speakers on their far from trivial efforts. I think a lot of conferences could learn a thing or two about appropriately expanding their list of topics and interests from EuroClojure.
What follows are some of my notes and impressions from a couple of selected talks — ones which hit close to home, or caused an "aha!" moment of sorts. For slightly more thorough notes, I strongly suggest you take a look at Philip Potter's incredible writeup.
The Talks
When I say "eclecticism", I do mean it: the whirlwind started with Fergal Byrne's talk on Hierarchical Temporal Memory, an intriguing approach to machine intelligence using state-of-the-art neuroscience research. I wasn't previously familiar with HTM, but the talk was pretty high-level, cumulating with an introduction to Fergal's Clortex, an equivalent of NuPIC in Clojure. His description of software development as more of a science rather than engineering resonated pretty well.
Gary Crawford's talk on sentiment analysis of tweets was just what the doctor ordered — it's something we've been working on at Neutrino. His interesting approach uses Redis' bit operations for categorization and quick population counts, something I definitely plan on exploring further. Along with his theatrical delivery of the Leiningen quote, the mention of Datensparsamkeit and his plea for privacy awareness when collecting data turned out to be one of the more memorable moments of the conference.
The best thing about Idris is that it lets you make condescending remarks about Haskells type system - @ctford #euroclojure
— Erik Assum (@slipset) June 27, 2014
Chris Ford's talk on Idris was a surprising one. Although he barely mentioned Clojure, it was a great, succinct overview of Idris' type system motivated by the explanation of the Curry-Howard correspondence: types being propositions (or theorems if you'd like), and functions being proofs. A nice dash of advanced type systems at a dynamic language conference, with a refreshing approach not as focused on theory. Good to see Clojure devs open to Idris! :)
Day two was also a lot about Om, with both David Nolen's keynote, and Anna Pawlicka's example-packed talk on reactive data visualizations. Both talks finally made a couple of things about Om and React click, and ultimately made me play around with the libs over the weekend. First impressions are brilliant, and even in its alpha stages Om has a bunch of potential to significantly change the way we think and act about UIs (it was about time!).
The Bottom Line
I'm doing great injustice by stopping here because there's a lot to be said about each of the talks, but Philip's notes linked above, as well as the eventual videos of the talks, will probably be saying a lot more than I ever could. I'm definitely looking forward to tinkering with a lot of cool new things presented — it really is an exciting time to be interested in new languages & technologies.
Thanks to the organizers, thanks to the speakers, and thanks to everyone present. Until next year! :)