On Monday and Tuesday, tutorials were held. I decided to start off with a PostgreSQL tutorial by Josh Berkus (@fuzzychef, PostgreSQL Experts Inc., see picture below) titled “The Accidental DBA”. Takeaways form his session include: recommendation to use of XFS or Ext4 on Linux to store the database files, store the logfiles on a separate drive/volume, three methods for taking backups: pg_dump, PITR, filesystem snapshot, many specific recommendations to check a database’s health.

In the afternoon, I attended James Ward’s tutorial titled “6 Minute Apps! Build Your First Modern Web App” (@_JamesWard, TypeSafe, see picture below). He showed us to write web apps by pulling together jQuery, Bootstrap, and some CoffeeScript to build the Client. The Server side could be anything that talks HTTP, but this session specifically used Play Framework with Java and Scala to build JSON REST services.

The first day ended with the Ignite plenary, which consisted of 5-minute introductions/talks (see picture below).

On Tuesday morning, I attended the Solr Quick start tutorial by Erik Hatcher (@ErikHatcher, LucidWorks, see picture below). Apache Solr is a Lucene-based blazing fast, highly scalable search engine used in many applications. It is a great platform to create an inverse index to search large text corpora. He also mentioned that it integrates nicely with the Apache Project Tika (reads pdf, MS Office, … documents).

On Tuesday afternoon, it was time for the Apache Drill tutorial by Ted Dunning (@Ted_Dunning, MapR, see picture below) and Jacques Nadeau (MapR, see picture below the one of Ted Dunning). Apache Drill is a new type of massively parallel processing (MPP) framework that allows companies to federate NoSQL and other data storage technologies (including e.g. relational databases) in a single query interface. In my personal opinion, I think there’s still work to be done to bring the Drill project to the level of maturity of other projects such as Hive.

On Tuesday evening they opened the show floor in the huge exhibit hall. Of course, we find the usual suspects in the exhibit hall such as Fedora, WordPress, Drupal, DataStax, ...

One booth in particular stood out in terms of traffic: the tiny Ubuntu booth (see picture below)... announcing the Ubuntu Edge: A combination of an Android Phone and desktop replacement running Ubuntu (of course) in one device.

In the picture below you see Alan Pope (@popey, Canonical) holding the new Ubuntu Edge device. He had the important task of bringing the prototypes to the event. More details can be found on his blog: click here. Of course, the only right action was to order it right away... which I did.

Wednesday morning the conference opened with some very strong keynote presentations. First, we were welcomed by Sarah Novotny (Meteor Entertainment) and Edd Dumbill (@edd). The picture on top of this page shows both on stage on Wednesday. I really liked Jay Parikh’s (@jayparikh, Facebook, see picture below) keynote talk on the Open Compute Project. The latter focuses on efficiency, and sustainability in the development of infrastructure technologies.

He discussed in particular the infrastructure aspects of Facebook’s photosharing features. Below, you see a graph characterizing the access patterns to photos. 82% of traffic goes to 8% of the photos during the first three months. This has led Facebook to design data centers dedicated to the storage of these kinds of data (see the smaller L-shaped building in the bottom part of the picture below).

Next, Mark Hinkle (@mrhinkle, Citrix, see picture below) took the stage to talk about the community-building aspect of open source. Moreover, he asks for more “open source” projects, not just in software (citing the example of the Open Prosthetic Project by @kuniholm, see below).

Next, Carin Meier (@carinmeier, Neo, see picture below) entertained us by showing the use of Clojure to control robots (in particular the AR Drone).

Next, inBloom Chief Product Officer Sharren Bates (@sharrensharren, see picture below) provides an overview of inBloom open-source technologies and its mission of vastly improving K-12 personalized learning.

Finally, Mark R. Shuttleworth (Canonical Ltd., see picture below), South African billionaire Ubuntu Linux founder, presents the Ubuntu Edge smartphone.

After the plenary session, I attended Kathleen Ting (@kate_ting, Cloudera)’s breakout session titled “Building an impenetrable ZooKeeper”. ZooKeeper plays the role of coordinator of distributed applications (Hadoop). Next, Suresh Srinivas (@Suresh_m_s, Hortonworks, see picture below) presented his talk titled “HDFS - What is New and Future”. HDFS is the HaDoop File System. The current major release, Apache Hadoop 2.0 offers several significant HDFS improvements including new append-pipeline, federation, wire compatibility, NameNode high availability, and performance improvements.

Before answering detailed questions, he ended his presentation by refuting some myths (see picture below).

The organizers also did a good job in providing us with a variety of heathy meals with lots of vegetables. The concept of bringing together people with the same interests by using “themed” tables was also very interesting (see picture below).

Moreover, some booths offered services to relieve attendees from potential stress...

In the afternoon, I chose to attend the “Raspberry Pi Hacks” session by Ruth Suehle (@suehle, Red Hat, see picture below), Tom "spot" Callaway (@spotrh, Red Hat). The Raspberry Pi device is a barebone computer at an very low price (currently its B-model is priced at 35 USD). The presenters gave some nice examples of how to put this device to good use. Moreover, they talked about the hardware options as well as software distro’s (also mentioning the upcoming Fedora release for Raspberry Pi) to be used with the ARM-powered device. They also co-authored a book that will be published soon by O’Reilly on this topic. In the audience, some people also brought their devices with them. E.g. Brad Jasperson showed his Raspberry Pi doubling as an mp3 player (see picture below).

On the show floor, this geeky device was shown (powered by a Raspberry Pi) by Domain.com (@domaindotcom). It monitors twitter for a specific hashtag and then dispenses candy)... click here to purchase the device. See also this link for more information by another OSCON attendee.

Esri Inc. had a booth to present their geospatial development platform showing off with some really nice maps (here represented by Jeff Archer, @vee_dubb).

The Xen virtualization project booth is shown here:

Next, Mark Grover (@mark_grover, Cloudera) gave a talk titled “Getting Hadoop, Hive and HBase up and running in less than 15 minutes”. The main topic of his presentation was Apache Bigtop – a project aimed at developing packaging and tests within the Hadoop ecosystem. This was probably one of my most favorite talks of the entire OSCON event. He illustrated this with nice hands-on exercises.

Next, I attended Russell Branca (@ChewBranca, Cloudant)’s talk titled “Map Reduce All the Way Down”. He talked about MapReduce, Disco (Python for map and reduce), CouchDB, B-trees.

In between, there was always the opportunity to have interesting one-on-one sessions with authors, programmers, ... . Here, it’s a session in the expo hall with Tim O’Reilly (@timoreilly).

Thursday morning started with plenary sessions. In the picture below, Matthew McCullough
(@matthewmccull, GitHub) and Sarah Novotny (@sarahnovotny, Meteor Entertainment) welcomed the attendees.

On the show floor, I ran into an interesting booth: OW2, a European-based open source community for infrastructure software.

Laura Weidman Powers (@laurawp, @CODE2040) kicked off the list of speakers with a plea for more diversity in the open source community. However, the year 2040 is the year when people of color will be the majority in the United States, but minorities are vastly underrepresented in the most important part of the US economy, the tech sector. Hence, the code2040 initiative.

Next, Tom Preston-Werner (@mojombo, CEO of Github, see picture below) took the stage, followed by Eileen Evans (Vice President & Deputy General Counsel, Cloud & Open Source, HP) who gave a talk titled “Licensing Models and Building an Open Source Community”.

Open source is quietly helping shape our society, and one can see its impact on government through the work of Code for America. In the picture below, Michal Migurski (@michalmigurski, Code For America), Tim O'Reilly (@timoreilly, CEO of O'Reilly Media, Inc.) and Jared Smith (@jsmith, Bluehost) discuss the topic how open source can help shape government.

Finally, Leigh Heyman (@WHWeb, Executive Office of the President of the United States, see picture below) is in charge of the website WhiteHouse.gov. He talked about the pioneering role to openness (open source, open data).

Next, I decided to attend Peter Zaitsev (CEO, Percona Inc)’s session on “Database Performance: What Really Matters?”, followed by Phil Dibowitz (@ThePhilD, Production Engineer, Facebook, see picture below) on the topic of “Scaling systems configuration at Facebook: the paradigms, design, and software behind managing massive numbers of systems with open source and small teams” using Chef.

After lunch, I attended Christopher Clark (@frencil, @sparkfun, Director of Information Technology, SparkFun Electronics Inc., see picture below)‘s talk titled “Open Sourcing Hardware with GitHub”. More and more hardware designs are also shared in the public domain (e.g. using GitHub).

In the exhibit hall, there was also a few hardware vendors showing their new equipment including SSD powered Silicon Mechanics servers achieving 12 Gb/s throughput.

Next, John A. De Goes (@jdegoes, CEO/CTO/Founder, Precog, see picture below) gave a nice talk titled “Introduction to Quirrel & R for Dummies”. It included a nice side-by-side comparison of both languages.

Finally, I attended Kevin Kluge (Elasticsearch)’s talk on the Apache CloudStack project - an open source project to build IaaS (Infrastructure as a Service) clouds. This was a great talk on a great project!

Friday morning John Graham-Cumming (@jgrahamc, CloudFlare) looks back on 30 years of programming. His conclusion is that there’s nothing new since 1983. This was a very entertaining presentation.

Next, I attended a great session on Drupal (@Dries) presented by Cassandra Wolff (SoftLayer, an IBM company).

Next, the Frank Willison Memorial Award 2013 was presented to Anna Martelli Ravenscroft (see picture below). Congratulations. She looked really cool wearing Google Glass.

The keynote session ended with a talk by Robert Lefkowitz (@r0ml, CTO, Sharewave). Finally, Piers Cawley (Senior Programmer, Thermeon) even made us sing and dance...

Next, I attended the breakout session titled “Mobile HTML5: Real Life HTML5 Use Cases for Mobile” by Tomomi Imura (@girlie_mac, Nokia USA, see picture below). She nicely highlighted powerful features in HTML 5. Admittedly, she also emphasized the fact that few browsers adhere 100% to the new standard. More details can be found on her technical blog.

Next, it was database time again. Christophe Pettus (@xof, @PGExperts, Programmer at PostgreSQL Experts, Inc.) presented “PostgreSQL as a Non-SQL Database”. He convincingly showed that PostgreSQL is a good choice not just as a relational database. Because of the added NoSQL features and others it even outperforms other Big Data databases not just in speed, but also in disk footprint.

In sum, OSCON 2013 was an excellent conference: Perfect organization by O’Reilly staff, packed with useful sessions/keynotes at an excellent venue (Portland Convention Center, see picture below). At the bottom of this page you see yours truly (@dirkvandenpoel) with Downtown Portland as a background.