Comment 6 for bug 1395655

https://launchpad.net/~ci-train-ppa-service/+archive/ubuntu/landing-005/+build/6723146

After introducing my patch, the project builds successfully on Qt 5.4. Upstream stopped using GLdouble and switched to float instead, but there are still many places where GL_DOUBLE is used. So for now I applied the same Qt-operation of re-defining both GLdouble and GL_DOUBLE to their GLES float counterparts as it was before.