Yesterday was the last university day with in depth, 3 hour talks. Here we go.
## “Lambda programming lab Hands-on Labs” by Stuart Marks, Angela Caicedo, Simon Ritter (fabian)
Day 2 of the Devoxx started off with a lot of Lambdas and functional programming: Two talks and one hands-on lab, each of which being scheduled for three hours.
Trying things out is more fun than watching Powerpoint presentations, so I attended the hands-on lab, which was run by Stuart Marks, Angela Caicedo, and Simon Ritter.
Exercises were handed out on USB sticks, and after a short introduction, everybody started to work on it. The speakers were available for questions, in case somebody got stuck.
The exercises were a bunch of simple Unit tests to be fixed, which made it easy to focus on the tasks without having to deal with the boilerplate. Moreover, solutions to all exercises were provided from the beginning, so it was possible to skip tasks, and to spend more time on tasks that were particularly interesting. The session was also a good chance to get in touch with other participants and discuss each other’s solutions.
This was a very good start into day two. As a side effect, I now have a Java 8 JDK with a lot of interesting examples on my laptop.
Jim Weaver gave a general introduction, why that ‘internet of things’ gets
Angela then take over and presented the Java based OS for the Legos Mindstorem series EV3, “LeJOS”. All sensors were presented and the motors available included. Getting LeJOS on the EV3 is quite involved but should not take to long.
The, the Raspberry Pi and the hardware possibilities. The Devoxx signage is demoed. It was designed with SceneBuilder and the deployment process was shown, which was pretty easy. Next a ‘real’ use case for the Raspberry Pi was presented by Johan Voss, which is a card reader for checking in for cultural events. Tipp: Use threading with JavaFX when using with embedded devices in order to avoid blocking on IO.
After the break, Gary Collins showed a tons of physical devices, explaining their characteristics. In a sample demo, an not-yet-released version of Netbeans is used for doing the Pi deployment. The the Duke Pad is presented, which is a Pi attached to a Touchscreen in a Box. However, there were quite serious technical issues, so except the boot sequence nothing could be showed ;-(. It is supposed to have a Java FX UI with a handful applications.
The next part of the show presented some results from the Pi Hackathon. Thomas showed an Pi with an running Camel Container as micro ESB and an attached Arduino borad with a speaker. The communication is done via AMQP and uses Vinicius AMQP API. Vinicius himself demoed is alcohol tester, much fun here.
Finally Simon Ritter gave some insights how he uses a Raspberry Pi for accessing the telemetry data from his car. The key device is a ELM32, which connets to the OBD-II part and broadcast information via Bluetooth and/or WIFI.
All it was quite entertaining, a wild mixture of hardware hackery. Nice.
This was the last talk for me on day two at Devoxx and it was in one of the BOF rooms.
These are normal rooms with tables. The room was very full so I had to sit
on the floor. First I was a little bit skeptical about it, but then it turns out that
it is a really interesting talk.
SPDY is going to be a new version of HTTP and will be called HTTP 2.0.
It is developed by Google and the talk was from the guys who implemented it
They compared the web at the beginning with the web from today and they
make clear that there are a lot of more request by one site then in the early days.
The problem with that is, that every request has a huge overhead because of the headers
SPDY tries to fix this problem by wrapping the HTTP-Request into a SPDY-Request.
The HTTP-Request will be compressed and the Headers will only be transferred by the first
request. After the first request that there will be a full duplex connection
between the client and the server as communication channel.
One additional feature is that the server can push resources to the client. So the client
don’t have request every resource by itself.
For the implemented Web-Application this happen completely transparent. It will get
the pure HTTP-Request only. The only thing you have to do is, configure your webserver using a
SPDY-Connector. A lot of Web-Browser already can use the protocol, so you get better performance for free.
They showed an example and the improvement was impressive. I think this is something everybody
should try by itself. I am happy that I stood a little bit longer on the day at Devoxx to see this talk.