Meeting C++ trip report

November 22, 2016

Last week I attended my first ever C++ conference. It was the Meeting C++ conference in Berlin. The conference was very nicely planned with some good speakers and good topics. There were only a few minor details which was inconvinient. Below I will try to express some thoughts about the different talks I saw and about the conference in general.

The Keynote

The keynote this year was presented by Bjarne Stroustrup. I have seen Bjarne a few times before and it never gets boring. He talked about what his thoughts were when he created C++ and how he is trying to stay true to those thoughts going into newer versions of C++. As always he talked expressing intent - both for the readers and the compiler. He then showed what kind of features we can expect from C++17 and what he hopes will come with C++2z. I am always impressed at the questions he gets and how he answers them. He is getting all kind of weird and technical questions and he answers them all in a very calm manner. There was one question he didn’t want to answer though, which was about which languages he likes and dislikes.

All your cache is belong to us

I also saw a talk about caches by Timur Doumler. It was very good. He had some excellent points accompanied by good figures and thought-out experiments. I never thought the subject of caching could become fun and interesting, but Doumler actually nailed it. Unfortunately, this presentation did not allow for questions due to time being short. This was the one single talk I had a question for. He also showed snippets of code as to how you would write better code if you are interested in cache friendliness.

A million lines of code

Next on the schedule was Peter Bindels from Tomtom talking about their own software. He talked about how they had had problems with way too many lines of codes. He presented some guidelines for how to handle it and talked about a tool Tomtom had developed to show internal code dependencies. This tool had been very successfull in helping them make their code more clear and expressive. He mentioned that their tool could find code dependency cycles and how you did want to have these. I found this talk very good. Peter was very engaging and talked about practical experience which is easy to convert into your own work.

A little dash of functional

The last talk of the day was Phil Nash with his ‘Functional C++ for Fun and Profit’. I was a little vary when joining this talk. Sometimes when people talk about functional C++ it just becomes a garbage can of metaprogramming. Nash had none of this. His talk was concise with some very nice points from his practical experience. He showed how to express intent with functional programming and how to strive towards the idea of purity. He talked a little about some persistent data structure before showing how to make some nice functional programming with lambda expressions and std::optional. It was very intereting. This was easily the best talk of the day for me.

Party time!

At this point we got some food. It was OK - nothing too good. And there was no beer with the meal! However, after the meal there was a C++ quiz by conan.io. It was very fun to try. We did OK, I think we got 5-6th place or something. After the quiz, at 9:00 PM the beer finally arrived. I went around, drank some beer and talked to some interesting people! This was also one of the best things about the conference: just talking to people about stuff. Too bad we only had one evening to do that. But next years Meeting C++ with one more day of conference will be better!

Saturday morning functional programming

First thing on the schedule saturday was ‘Functional Reactive Programming in C++’ by Ivan Cukic. I have to give it to Cukic for bringing a very relevant and easy to use presentation! Again, I had expected some metaprogramming, but it was good on the conceptual level as to how you want to explore funcation reactive programming! Was an excellent presentation. Cukic also shared free copies of his future book on this exact subject!

Static analysis!

Next up was Gabor Horvath with his talk about static analysis. Gabor is a PhD student with subject in static analysis, so this talk was a little academic. But it was indeed very interesting. He had good examples and nice slides on the subject. When the presentation was over I was much more curious about static analysis and wanted to look more into it. This was also one of my favorite talks of the conference.

The end

Unfortunately I didn’t get to see the closing keynote due to my travel plans. I had to get home. But that was OK - I had had an awesome first conference. I am very inspired about the things I learned there. I can feel there are a lot of things I still need to learn and I look forward to this! I got to speak with some interesting people and share laughs with others! I hope I can take some of the subjects I heard about and incorporate them into my work.

This was all in all a good conference and an awesome experience!

Discussion, links, and tweets

I'm a developer and CS graduate. Follow me on Twitter; you'll enjoy my tweets.

I also keep an "ask me anything" type of project in a repository on GitHub (naturally!). Feel free to ask me a question.