Thursday, 03 May 2012
I'm now back home again from ACCU 2012. As ever, it was an exhausting, but thoroughly enjoyable conference. With 5 tracks it was sometimes very hard to choose which session to attend.
There were a surprising number of talks on C, though mostly just to confirm that C is a dying language, a "zombie language" as Uncle Bob put it, even if there is still a niche or two where it is the best fit.
There was also a fair number of sessions on testing, especially TDD (even TDD in C), and a few sessions on other agile practices.
However, the big topic of the conference was C++11. Nicolai Josuttis ran a full day pre-conference tutorial on the C++11 library, and then there were sessions on the best and worst features of the C++11 library, C++11 for everybody, C++11 for the rest of us, C++11 allocators, writing generic code with C++11, the C++11 concurrency facilities, an interactive C++11 programming session, and even a C++11 pub quiz (with free beer!)
The other big topic was concurrency and parallelism, with 6 presentations, including two by me. Actors and Dataflow seemed to be a common theme — it looks like these ideas are finally filtering into the mainstream now that everyone and his dog has a multicore computer in their pocket. Not only did Russel, Jason and I reference them in our presentations, but several of the developers I spoke to had implemented their own actor or dataflow libraries, or were interested in using one.
My own presentations went well. My first one (on Dataflow, Actors and High Level Structures in Concurrent Applications, on Thursday afternoon) was packed out. There were quite a few interesting questions from the audience, and someone told me it was the best presentation at the conference, which was really nice to hear. Several people asked about slides and the code samples, and I promised to make them available. The slides are here and the code samples here
My second presentation was the C++11 concurrency tutorial on Saturday morning. I was presenting in the first slot after the keynotes, and the speakers dinner was the night before, so a lot of people had been looking rather the worse for wear in the beginning. However, the room was almost full again, and the audience seemed to be awake — yet again there were many interesting questions that showed people were paying attention. The slides for are here and the code samples here
My book was only published a couple of months before the conference, so it was exciting to see so many people clutching copies. Apparently the Blackwell's stall sold out, which was nice for me, but unfortunate for those that didn't manage to get a copy. If you wanted a copy then you can buy it direct from Manning, or from amazon.com, or amazon.co.uk.