We’ve so far had a slow start with our thesis. There has been some labs and reports that needed to be written. But there has been some progress!
I’ve taken the role of being in charge of making sure that everything works for everybody, that entails using cmake since we run multiple platforms. The first step was to work with dependencies. I started trying to use git submodules for this, but it ended up being too much of a hassle due to dependencies of dependencies. OpenCV was quite simple to get working this way, but PCL and the kinect library was trickier. So I decided to just have a race of having everyone install the need libraries. I’ll have another try with submodules and cmake at a later stage to simplify the delivery to our client. Since everything is statical linked, a distribution of the system would work but the client wanted to be able to build it by himself.
Our teacher had some lecturers where he talked about continuous integration and build servers. It got me interested in the use of a build server. I ended up trying out Jenkins - And it’s pretty awesome. So far I have automatic builds on github pushes, automatic generation of doxygen and publishing it. I’ve also fixed static code analysis and google test reports on every build. If anything wouldn’t work - the user that pushed the code to github will get an email saying something is wrong.
In the future I also want to look at cross-compling and analysis of test coverage.
The real work will begin tomorrow with the start of our first sprint!