Random Thoughts form Øredev 2010
November 14, 2010
Last week I was at the Øredev 2010 conference and it was the first IT conference I have attended. I have, however, attended a wastewater conference around 2005, but that still leaves me little to compare Øredev to.
Monday I attended an all-day seminar on REST, which I found to be quite interesting, even if I left feeling a little confused. The two presenters, Jim Webber and Ian Robinson, were good and have written the book “REST in Practice”, which I will properly buy (and yes, I got a discount ticket :) ).
I worked on Thursday, as this conference day did not look particular interesting. I will not go into all the presentations I attended, but will say I few words about the most interesting ones.
Giorgio Sardo from Microsoft did the talk “Deep Dive into HTML5”, where here presented HTML 5 in general and Microsoft’s HTML 5 work in particular. He did some live coding, which worked out really well. I am, however, concerned about HTML 5, as it seems extraordinary complex. The map of HTML 5 may give you an idea – if you like me, don’t have Silverlight, then look at the picture. Each box in the picture is a standard, which is part of HTML 5 and thus leaves room for specification errors and implementation errors by vendors. I just cannot see have such a huge set of specs can ever be implemented without gigantic headaches for implementors as well as users. Granted, I am taking the 5-minute 12 kilometer view – but still, color me skeptical.
Simultaneously with the HTML 5 talk, Jonas Bonér did a talk about Akka, which is a concurrency framework for Scala. Jonas is a good presenter and has made what seems (never tried it) a nice piece of technology. But a half year ago, I saw him make a two hour presentation of Akka, and thus saw no point in attending this one hour talk.
Later Wednesday, I attended the talk “Robotium Testing for Android” by Hugo Josefson and Renas Reda. About a week before the conference, I had started using Robotium myself and it is really nice. Furthermore, the vanilla Android testing API just sucks, as better explained in this blog post over at Brain Flush.
Thursday, Sean Cribbs talked about Webmachine, which is a quite novel HTTP implementation. The main selling point, is that it leaves a large degree of control in the hands of the Webmachine user (a programmer). Webmachine also has a very cool debugging tool, where you can see a flow diagram of a HTTP response process and what happened at each step. Webmachine is intended to be used for REST applications, where more fine grained control of HTTP may be needed. Very nice. Webmachine is implemented in Erlang.
Marcus Lagergren and Marcus Hirt presented the JRocket JVM Friday. They managed to give a good feel for the implementation of an adaptive JVM, in just an hour, which is quite impressive. They also showed some very nice debugging tools they were working on.
All in all, the conference was a nice experience. However, there were a few bad things as well. First of all, too many of the talks had nothing to say or talked without really getting to any points. I think part of the problem, was that I could have selected the talks to go to better – better planing I guess. I was half-sick Wednesday though Friday, so I was not motivated to investigate the different speakers and Øredev did not make it easy. It would have been really great, if Øredev had the speakers slides online, so that one could have had a better idea about what a speaker would say.
On a final note, I was disappointed about the food and the physical environment. Some rooms were quite cold and the noise level from talks “next door” were often annoyingly laud. From other attendees, I got the clear impression that Øredev had better food and as good rooms as other IT conferences, thus it is properly just me who has been spoiled by a wastewater conference, where the food and rooms were exceptionally good.